<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" xmlns:googleplay="http://www.google.com/schemas/play-podcasts/1.0"><channel><title><![CDATA[Product: Behind the Craft]]></title><description><![CDATA[Real lived stories from product leaders, for product leaders and aspiring leaders. The issues they faced, the lessons they learned, and how you can apply them in your own day-to-day in product.]]></description><link>https://stories.logrocket.com</link><image><url>https://substackcdn.com/image/fetch/$s_!CKg4!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F41670c83-3afd-46d0-91fe-e11d75bfe508_600x600.png</url><title>Product: Behind the Craft</title><link>https://stories.logrocket.com</link></image><generator>Substack</generator><lastBuildDate>Sun, 22 Nov 2026 09:55:48 GMT</lastBuildDate><atom:link href="https://stories.logrocket.com/feed" rel="self" type="application/rss+xml"/><copyright><![CDATA[LogRocket]]></copyright><language><![CDATA[en]]></language><webMaster><![CDATA[productbehindthecraft@substack.com]]></webMaster><itunes:owner><itunes:email><![CDATA[productbehindthecraft@substack.com]]></itunes:email><itunes:name><![CDATA[Jeff Wharton]]></itunes:name></itunes:owner><itunes:author><![CDATA[Jeff Wharton]]></itunes:author><googleplay:owner><![CDATA[productbehindthecraft@substack.com]]></googleplay:owner><googleplay:email><![CDATA[productbehindthecraft@substack.com]]></googleplay:email><googleplay:author><![CDATA[Jeff Wharton]]></googleplay:author><itunes:block><![CDATA[Yes]]></itunes:block><item><title><![CDATA[Why the Best AI Bet Is a Flywheel, Not a Feature | Rajiv Jain, SVP Product Strategy (AliveCor)]]></title><description><![CDATA[Rajiv Jain explains why deployed models create their own data, how to build a roadmap backward from lives saved, and what building Alexa taught him about AI agents and business models.]]></description><link>https://stories.logrocket.com/p/why-best-ai-bet-flywheel-not-feature-rajiv-jain</link><guid isPermaLink="false">https://stories.logrocket.com/p/why-best-ai-bet-flywheel-not-feature-rajiv-jain</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 17 Nov 2026 14:46:19 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/8d2d754e-1c5e-464e-9649-4fd5aaa0874d_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p></p><div id="youtube2-KtKCjoxCjHI" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;KtKCjoxCjHI&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/KtKCjoxCjHI?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI">YouTube</a> | <a href="https://open.spotify.com/episode/2KtitXjLGjvgIAv7RLZrEZ">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/why-the-best-ai-bet-is-a-flywheel-not-a/id1733103005?i=1000802476399">Apple</a></strong></em></p></div><p>It seems like AI lets anyone ship anything. So where&#8217;s the moat? Our guest says it&#8217;s not the model &#8212; it&#8217;s what happens after you ship that matters most.</p><p>Rajiv Jain is Senior VP of Global Product Strategy at AliveCor, maker of the personal ECG that gives you a 30-second, AI-powered heart reading at home. Six million users later, every recording makes the next model smarter. That&#8217;s how AliveCor went from detecting one heart condition to six, with 26 on the roadmap. </p><p>Before that, Rajiv built Amazon&#8217;s Alexa enterprise ecosystem, including work with NASA on an AI companion for the trip to Mars, and a test run on Artemis I around the Moon.</p><p>In this episode, Rajiv shares:</p><ul><li><p>Why the smartest bet in AI is a flywheel, not a feature</p></li><li><p>How to build a roadmap backward from lives saved, then work out the tech</p></li><li><p>And what building Alexa taught him about AI agents, and why the business model matters more than the tech</p></li></ul><p>So, here&#8217;s our episode with Rajiv Jain.</p><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. Deploying a model creates its own data</h2><p>AliveCor has a big proprietary dataset, and it also buys data from other sources. But Rajiv argues the most valuable data comes from shipping the model in the first place.</p><p>When a patient uses an AliveCor device, they don&#8217;t record once. They might take 18 to 28 recordings a month, and keep recording for years. That turns a single snapshot into a longitudinal record.</p><blockquote><p>&#8220;So I have not just one point-in-time data, but data over a period of time. In that, I also know what&#8217;s a patient demographic, what conditions are they developing over a period of time.&#8221;</p></blockquote><p>And patients add more than heart readings. They log how they&#8217;re feeling, what they ate, and how active they were. Over time, the model&#8217;s inputs go from a single signal to many. That&#8217;s what gets each generation of model better than the last:</p><blockquote><p>&#8220;Deploying that model creates a flywheel, and then generations of model get better and better and better just because you&#8217;re getting more and more types of signals into the model.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Don&#8217;t wait for the perfect dataset before you ship. Your proprietary data can get you started, but the data your product generates once it&#8217;s in customers&#8217; hands is what will compound.</p><div><hr></div><h2>2. Start with lives saved, then work out the tech</h2><p>Because AliveCor makes its own hardware, Rajiv thinks of its devices as sensors. Add more sensors, and you can capture more data. But what to build next doesn&#8217;t start with the data. It starts with the value.</p><p>Rajiv gives an example. A doctor in a hospital might use an ECG to detect about 100 conditions. New research suggests the ECG signal could detect up to 600. Some of those conditions currently need an echocardiogram, which can cost hundreds or thousands of dollars and take time to schedule.</p><blockquote><p>&#8220;If you can do a low-cost ECG just to be able to detect that condition that you used echo for, then all of a sudden we have reduced friction.&#8221;</p></blockquote><p>Now imagine bringing that detection into the patient&#8217;s home. That&#8217;s the kind of opportunity AliveCor&#8217;s roadmap is built around. From there, the team works with doctors to correlate hospital data and patient records with the ECG data AliveCor has gathered over the years, to see whether it can detect a condition reliably.</p><p>Only then does the team ask what it can build with the data and signals it has, and what it can take through the FDA and other regulators.</p><p><strong>Product takeaway:</strong> Build your roadmap backward. Start by quantifying the value for each audience you serve: in AliveCor&#8217;s case, lives saved, time to treatment, and cost. Then work out whether your data and tech can deliver it. It keeps the team focused on the outcomes that matter instead of chasing whatever the newest model makes possible.</p><div><hr></div><h2>3. Get every layer of the system to see the value</h2><p>Plenty of great healthcare products fail because the incentives don&#8217;t line up. Patients, providers, and payers all need to see a benefit, or the best tech in the world goes nowhere.</p><p>AliveCor started direct to consumer. Patients buy the device out of pocket, so the first job was to serve them well when they&#8217;re not in front of a doctor. But the team quickly realized it was also serving the doctors working with those patients.</p><blockquote><p>&#8220;50 to 60% of my patient journeys start in a doctor&#8217;s office. Because over a period of time, doctors realize that, hey, if the patient&#8217;s more educated and they are more informed about their health and they&#8217;re empowered, they&#8217;ll respond to treatment better.&#8221;</p></blockquote><p>Once value is proven with patients and doctors, AliveCor can take that case to payers and employers, in the US and in Europe, covering the financial, medical, and scientific sides of the conversation.</p><p><strong>Product takeaway:</strong> In a market with many stakeholders, you don&#8217;t have to win everyone at once. Pick the one you can serve directly, prove the value there, and use that proof to bring in the next layer. Each win builds the case for the next one.</p><div><hr></div><h2>4. What Alexa taught Rajiv about AI agents</h2><p>Rajiv joined Alexa around 2016, when voice assistants were just starting to capture people&#8217;s imaginations. His job was to take Alexa into the enterprise.</p><p>One of the most ambitious projects was with NASA. A trip to Mars could keep astronauts locked in a small space for months or even years. How do you keep them sane, engaged, and productive? Maybe Alexa could help.</p><blockquote><p>&#8220;None of us wanted the HAL, open the pod door scenarios. But really thinking about, &#8216;Hey, what are the mental aspects? What are the challenges an astronaut is going to face?&#8217;&#8221;</p></blockquote><p>That work led to experiments with NASA, including a test run of Alexa on the Artemis I mission around the Moon. Rajiv&#8217;s team also worked with Disney on Disney-branded devices in hotel rooms, where guests could check reservations and showtimes, and a Disney skill that brought some of the park experience home.</p><p><strong>Product takeaway:</strong> The use cases for AI agents haven&#8217;t changed much in a decade. What&#8217;s changed is how well the tech can deliver them. Before you start building, look at what earlier waves of assistants tried and where they fell short. And plan for your agent working alongside other agents, with clean handoffs, rather than as a standalone experience.</p><div><hr></div><h2>5. Invest in flywheels, not features.</h2><p>Rajiv&#8217;s hardest-won lesson from Alexa is that great tech isn&#8217;t enough on its own:</p><blockquote><p>&#8220;Technology is important, customer experiences are important, but the business model and the financials are equally important.&#8221;</p></blockquote><p>Amazon could sell the devices, but how would the usage make money? With today&#8217;s inference costs, that question matters even more.</p><p>His answer is to invest in flywheels: a better model brings more data, which reaches more customers, which makes the model better again, over five or six years.</p><blockquote><p>&#8220;Don&#8217;t look at a product, don&#8217;t look at a feature. Really think about if there&#8217;s a flywheel behind it, what does that flywheel look like and how do you make sure you&#8217;re making the right investments.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Start every bet with a hypothesis about what customers will pay for and how the economics hold up. Then back the investments that feed a flywheel for years, not just a feature for this quarter.</p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI">00:00</a> Introduction<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI&amp;t=223s">03:43</a> How AliveCor's at-home ECG uses AI to read your heart<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI&amp;t=407s">06:47</a> How deployed models create a healthcare data flywheel<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI&amp;t=651s">10:51</a> Prioritizing features by lives saved<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI&amp;t=804s">13:24</a> Why AliveCor started direct to consumer, then won over doctors and payers<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI&amp;t=954s">15:54</a> What building Alexa taught Rajiv about AI companions<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI&amp;t=1179s">19:39</a> Bringing Alexa to Disney hotel rooms and Disney characters into homes<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI&amp;t=1277s">21:17</a> Why the business model matters as much as the AI itself<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI&amp;t=1497s&amp;pp=0gcJCWMAwfN6Pr3D">24:57</a> Alexa's monetization problem and why to invest in five-year flywheels<br><a href="https://www.youtube.com/watch?v=KtKCjoxCjHI&amp;t=1657s">27:37</a> Conclusion</p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/rajivjain/">Rajiv&#8217;s LinkedIn</a></p></li><li><p><a href="https://alivecor.com/">AliveCor</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/">LogRocket.com</a>.</p><p></p>]]></content:encoded></item><item><title><![CDATA[AI Made Us 4x Faster. Now Humans Are the Speed Limit | Sharon Hunt, JLL]]></title><description><![CDATA[Sharon Hunt talks about why AI stalled on her team's older codebase at first, what it took to fix that, and the new problem that showed up once the team was moving four times faster.]]></description><link>https://stories.logrocket.com/p/ai-made-us-4x-faster-user-adoption-new-speed-limit-now-sharon-hunt</link><guid isPermaLink="false">https://stories.logrocket.com/p/ai-made-us-4x-faster-user-adoption-new-speed-limit-now-sharon-hunt</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 10 Nov 2026 14:05:37 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/7b3b299a-29a4-4570-b9cc-7b986f78dc03_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p></p><div id="youtube2-8_QWBOG7FDQ" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;8_QWBOG7FDQ&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/8_QWBOG7FDQ?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ">YouTube</a> | <a href="https://open.spotify.com/episode/7D4HudomkToYbCqD9ausQy">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/ai-made-us-4x-faster-user-adoption-is-the-new-speed/id1733103005?i=1000800997608">Apple</a></strong></em></p></div><p>You&#8217;ve probably heard that building software is basically free now. </p><p>That&#8217;s only true if you&#8217;re starting from scratch. It&#8217;s a different story at a company like JLL, a commercial real estate firm that&#8217;s been around for about 200 years.</p><p>Sharon Hunt leads product for JLL&#8217;s property management technology. That includes Building Engines, a company JLL acquired. Its software runs the day-to-day of commercial buildings: work orders, visitor access, inspections, and tenant requests. JLL uses it in every building it manages and also sells it to other property management firms.</p><p>When Sharon&#8217;s team first tried AI coding tools on their main product, the results weren&#8217;t usable. So they spent a year getting the product &#8220;AI-ready,&#8221; and now they ship about 4x faster. But that created a new problem: the people around the product can&#8217;t keep up with how fast it changes.</p><p>In this episode, we get into:</p><ul><li><p>Why AI struggled with an older codebase, and the cleanup that made it 4x faster</p></li><li><p>Why the new bottleneck is how much change your sales team, support team, and customers can absorb</p></li><li><p>How clear customer metrics got engineers bringing their own product ideas</p></li><li><p>When to ship small fixes right away and when to hold changes for one big release</p></li></ul><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. Slow down to speed up: Getting an older product ready for AI</h2><p>Building Engines&#8217; main product, Prism, was rewritten in 2018. That&#8217;s recent for a company founded in 2001, but eight years isn&#8217;t exactly <em>new</em> for software. Parts of the code had little documentation, logic was nested several layers deep, and the front end was mostly custom-built with no shared design system.</p><p>When the team pointed AI coding tools at it, the AI got just as lost as a new hire would.</p><blockquote><p>&#8220;What we learned trying to deploy, and speed up development on this application using Claude early on, was that Claude was running into a lot of the same challenges that our new engineers were running into.&#8221;</p></blockquote><p>The team decided to clean up first. They started small. A team of three or four rebuilt the tenant portal, a lower-risk part of the product that tenants only use now and then. It took about a quarter instead of the year they&#8217;d estimated. Along the way, they built reusable AI workflows and adopted Solstice, a JLL design system built to work well with AI tools. Then they used that playbook to help a 70-person R&amp;D team start rewriting the main app for property managers.</p><p>That rewrite is now about 80% done. The team works through about four times as much as before, and cleared more than 100 backlog requests along the way. The cost was real: they mostly stopped shipping major new features for about six months to get there.</p><p><strong>Product takeaway:</strong> If AI tools are struggling in your codebase, they may be hitting the same problems your new hires do. Fixing those problems (clear structure, documentation, a shared design system) pays off for both. Test the approach on a low-risk part of the product first, then use what you learn to scale it.</p><div><hr></div><h2>2. The new speed limit is how much change people can absorb</h2><p>Once the engineering team got fast, the bottleneck moved somewhere else. Every change has to be learned by sales, so they can sell it accurately. Customer success has to learn it so they can answer questions, and implementation teams have to learn it so they can onboard new buildings.</p><p>Then there are the users. Property managers use the product all day to do their jobs, so constant big changes wear them out. Their tenants are also customers, and if tenants get frustrated, that frustration comes back to the property team and then back to JLL.</p><blockquote><p>&#8220;It&#8217;s all of the scaffolding and the enablement around the product...it&#8217;s human absorption that&#8217;s the new limiting factor.&#8221;</p></blockquote><p>Sharon would still rather have this problem than the old one. Being too fast for your organization beats being too slow for your market.</p><p><strong>Product takeaway:</strong> When AI speeds up your engineering team, check what slows down next. Plan for how quickly sales, support, and customers can learn what you ship, and treat that capacity as a real constraint on your roadmap, not something to deal with after launch.</p><div><hr></div><h2>3. Give the team a &#8220;why,&#8221; and they&#8217;ll find the ideas</h2><p>Before the rebuild, the roadmap was full of urgent, disconnected requests: features for new countries, and gaps to close so clients moving from older Building Engines products would sign. None of it added up to a strategy.</p><p>Sharon replaced that with three measurable benefits for customers:</p><ul><li><p><strong>Saving property staff time</strong> on repetitive work, so they can spend it on tenant relationships</p></li><li><p><strong>Increasing net operating income</strong> by finding ways the product can raise revenue or cut costs for a building</p></li><li><p><strong>Faster deployment</strong>, so a new building or account can go live quickly</p></li></ul><p>Every feature request now gets checked against those three. How many people use this workflow? How much time would the change save? The result she&#8217;s most excited about:</p><blockquote><p>&#8220;I&#8217;m seeing for the first time a lot of really great ideas coming back from the R&amp;D teams, specifically developers, because now they understand these metrics.&#8221;</p></blockquote><p>Engineers also have more time to explore the product, spot any issues, and fix them on their own. And when teams visit buildings to watch people work, they can come back with a working prototype the same day. That used to take weeks or months.</p><p><strong>Product takeaway:</strong> &#8220;Build this by April 1st&#8221; gets you a feature. &#8220;Here&#8217;s the problem and why it matters&#8221; gets you ideas. Define a few customer benefits you can measure, share them with the whole team, and use the time AI gives back to get engineers closer to customers.</p><div><hr></div><h2>4. Ship small fixes now, save big changes for one release</h2><p>If your team can ship 100 new features in a day, most customers will never find them. Sharon&#8217;s approach is to make the workflows people already use better before adding new ones. Take certificates of insurance: property staff used to type in every field by hand. Now they can drop in a file and let the product fill it in, all inside a step they already know.</p><p>For releases, she splits changes into two groups:</p><ul><li><p><strong>Small fixes ship right away.</strong> Things like letting users select all in a dropdown don&#8217;t need training.</p></li><li><p><strong>Big workflow changes wait for one release.</strong> Her team is rethinking inspections so an engineer can just take photos, record a video, or dictate notes instead of filling out a questionnaire. That changes the reports, the mobile app, and everything downstream, so it all ships together.</p></li></ul><blockquote><p>&#8220;We&#8217;ll be making so many changes, you want to be choiceful with what you ask your customers to absorb.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Sort what you ship by how much it asks users to relearn. Ship small fixes continuously. Bundle changes that alter how a workflow works into one release, with training ready for customers, sales, and support at the same time. Now that building is cheaper, rethink whether a small first version is still the right call.</p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ">00:00</a> Introduction<br><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ&amp;t=248s">04:08</a> Why AI struggles with an aging codebase<br><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ&amp;t=316s">05:16</a> Making JLL's Prism AI-ready by rebuilding the tenant portal in one quarter<br><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ&amp;t=543s">09:03</a> AI lets you ship 4x faster, but users can only absorb so much change<br><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ&amp;t=727s">12:07</a> JLL paused major features for six months to go slow to go fast<br><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ&amp;t=1197s&amp;pp=0gcJCWMAwfN6Pr3D">19:57</a> Improving existing workflows instead of building new features<br><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ&amp;t=1339s&amp;pp=0gcJCWMAwfN6Pr3D">22:19</a> Finding friction with data, site visits, and engineers&#8217; own ideas<br><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ&amp;t=1642s">27:22</a> When to ship small tweaks and when to hold back for big releases<br><a href="https://www.youtube.com/watch?v=8_QWBOG7FDQ&amp;t=1908s">31:48</a> Conclusion</p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/sharonehunt/">LinkedIn</a></p></li><li><p><a href="https://www.jll.com/en-us/">JLL</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[AI Can Design Now. Does UX Still Need a Human? | Mykola Konrad, CPO at Euna Solutions]]></title><description><![CDATA[Euna Solutions' CPO, Mykola Konrad, on which UX work is already moving to PMs, developers, and agents, which work stays human, and how to get your team through the change before it gets through you.]]></description><link>https://stories.logrocket.com/p/ai-can-design-now-does-ux-still-need-human-mykola-konrad</link><guid isPermaLink="false">https://stories.logrocket.com/p/ai-can-design-now-does-ux-still-need-human-mykola-konrad</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 03 Nov 2026 14:21:21 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/83b56dee-37c7-4ea7-bd7f-04736ece4bbf_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div id="youtube2-fBQSC_k9ZbI" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;fBQSC_k9ZbI&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/fBQSC_k9ZbI?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI">YouTube</a> | <a href="https://open.spotify.com/episode/0tDphhgmc03Pz8oe0ZBRBI">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/ai-can-design-now-does-ux-still-need-a-human-mykola/id1733103005?i=1000799502113">Apple</a></strong></em></p></div><p>A month ago, someone on Mykola Konrad&#8217;s team was still building prototypes by hand. Today, an agent does it.</p><p>That&#8217;s how fast the ground is moving under UX teams right now. And when we ask product leaders where design is headed, nobody agrees. Is it all agents from here? Does AI design everything?</p><p>Mykola has a more practical answer than most. He&#8217;s CPO at Euna Solutions, which builds software for government agencies. Before that, he spent 25+ years in product at companies like Microsoft, Avaya, and Brightcove. At Euna, UX reports directly to him, so the question of what happens to design isn&#8217;t theoretical. He&#8217;s making those calls right now.</p><p>In this episode, we get into:</p><ul><li><p>Which UX work is moving to PMs, developers, and agents, and the &#8220;If you can, then do&#8221; mantra Euna is using to get there</p></li><li><p>Why customer contact, high-stakes flows, and accessibility still need a human making the final call</p></li><li><p>Myk&#8217;s change management playbook for getting a team through AI adoption, and why token spend is a useful (if imperfect) signal</p></li><li><p>How Euna is running two delivery flows side by side, AI-enhanced and full-AI, and why customer adoption may be the real speed limit</p></li></ul><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. The UX work that&#8217;s already succeeding: &#8220;If you can, then do&#8221;</h2><p>Myk doesn&#8217;t think all of UX is up for grabs. But some of it is moving out of the UX team</p><p>The work on its way out is the work that doesn&#8217;t need deep taste or customer context: design system upkeep, adding new components, first drafts, early exploration, and getting from a PRD to a prototype. That work still happens. It just happens somewhere else.</p><blockquote><p>&#8220;The activity still happens, but it will move to a different persona within the organization. And right now we are actually having some product managers do that, and we&#8217;re also having some developers do that, &#8216;cause I&#8217;m not sure exactly 100% where it&#8217;s gonna land.&#8221;</p></blockquote><p>Rather than deciding up front who should own it, Euna is letting people volunteer:</p><blockquote><p>&#8220;If you as an individual believe you have the aptitude or the skill set or just the desire to try to prototype something using Claude Code or Claude Design, then do it. Don&#8217;t wait for the UI/UX team to help you out, just try it.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> You don&#8217;t need to know exactly where every task will land before you let people start. Give anyone with the aptitude permission to pick up the lower-stakes design work now, watch who takes to it, and use what you learn to redraw the roles later. Just be upfront that &#8220;optional&#8221; won&#8217;t stay optional forever.</p><div><hr></div><h2>2. What stays human?</h2><p>So what doesn&#8217;t move? For Myk, it comes down to three things.</p><p><strong>1. Direct customer contact.</strong> Myk has been interviewed by an AI agent himself, and it was fine, up to a point.</p><p>For some of Euna&#8217;s customers, it&#8217;s not even a matter of quality. Several large state and provincial governments simply won&#8217;t work with an agent:</p><blockquote><p>&#8220;They will not accept an agent interfacing with them. I wouldn&#8217;t even call it regulatory, it&#8217;s partly the way that that particular government entity needs and wants to operate. They want a human being doing it.&#8221;</p></blockquote><p><strong>2. The final call on high-stakes flows.</strong> Euna owns the rails on its payments product, and it builds compliance flows and audit trails. AI can give you a first pass, but someone with deep UX expertise needs to decide what actually ships.</p><p><strong>3. Accessibility and regulated work.</strong> Even as AI improves and takes on more interviews and insight work, Myk expects accessibility and compliance to keep a human in the loop.</p><p><strong>Product takeaway:</strong> Sort your UX work by the cost of getting it wrong, not by how hard it is to do. Low-stakes, easy-to-review work is where AI should go first. Anything involving money moving, regulators, or users with accessibility needs should keep a human making the final call (even if the AI draft looks good).</p><div><hr></div><h2>3. Getting your team through the change</h2><p>How do you get an existing team ready for a future that looks this different? Myk starts with change management 101: almost nobody likes change. Some people just get through the skeptical phase faster than others.</p><p>From there, Myk&#8217;s approach is solid:</p><ul><li><p><strong>Pick the tools: </strong>Euna&#8217;s design team is a two-tool shop right now: Figma Make plus Claude Design for prototyping. Everyone gets trained on both</p></li><li><p><strong>Map the new handoff: </strong>How does work move from design agents to product to the &#8220;software factory&#8221;?</p></li><li><p><strong>Reset expectations on time: </strong>If a design used to take three weeks, try three days and see what happens</p></li><li><p><strong>Watch for signals: </strong>Myk wasn&#8217;t trying to monitor anyone, but token spend made it clear who was jumping in and who wasn&#8217;t. It&#8217;s not a productivity metric on its own, but someone using almost nothing probably isn&#8217;t changing</p></li></ul><p>He also recommends asking your team directly where the bottleneck is. When Euna&#8217;s product and dev teams went agentic, the answer that came back was UX. From there, the question is whether it&#8217;s a data-in problem (user research isn&#8217;t done because customers haven&#8217;t responded yet) or a data-out problem (the research exists, but designs aren&#8217;t moving fast enough).</p><p><strong>Product takeaway:</strong> Be honest about your timeline before you design your change plan. The length of the window you can give people should shape how much training, patience, and support you build in. And don&#8217;t guess where the slowdown is: ask the team, then figure out whether it&#8217;s a process gap or a data gap before you try to fix it.</p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI">00:00</a> Introduction<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI&amp;t=140s">02:20</a> Myk's path from electrical engineer to CPO<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI&amp;t=201s&amp;pp=0gcJCWMAwfN6Pr3D">03:21</a> Will AI replace UX?<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI&amp;t=307s">05:07</a> UX work shifting to PMs, developers, and agents<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI&amp;t=395s">06:35</a> Where UX still needs humans: Customer research, payments, and compliance<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI&amp;t=827s">13:47</a> Myk's change management playbook and token spend as an adoption signal<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI&amp;t=1040s">17:20</a> Finding bottlenecks in an agentic delivery process<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI&amp;t=1145s">19:05</a> AI-enhanced vs. Full-AI workflows from product to dev<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI&amp;t=1471s">24:31</a> Customer adoption is the real speed limit<br><a href="https://www.youtube.com/watch?v=fBQSC_k9ZbI&amp;t=1543s">25:43</a> Conclusion</p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/mykolakonrad/">Myk&#8217;s LinkedIn</a></p></li><li><p><a href="https://eunasolutions.com/">Euna Solutions</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[Leader Spotlight: What makes large-scale transformations stick, with Hugh Theodore]]></title><description><![CDATA[Hugh Theodore is Senior Director of Product Management at Dayforce, a global leader in HCM technology.]]></description><link>https://stories.logrocket.com/p/leader-spotlight-hugh-theodore</link><guid isPermaLink="false">https://stories.logrocket.com/p/leader-spotlight-hugh-theodore</guid><dc:creator><![CDATA[Jessica Srinivas]]></dc:creator><pubDate>Thu, 29 Oct 2026 07:02:57 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!nIqf!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><em>Hugh Theodore is Senior Director of Product Management at Dayforce, a global leader in HCM technology. Hugh focuses on the company&#8217;s tax and payments organization. Since joining Dayforce in 2021, he has also served as Director, Product Readiness and as Senior Director, Product Operations. Prior to Dayforce, Hugh held product leadership roles at Sensibill, Yuhu, and Symbility Intersect. Earlier, he worked in implementation and consulting at mindSCOPE Staffing Software and Konrad Group.</em></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!nIqf!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!nIqf!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!nIqf!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!nIqf!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!nIqf!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!nIqf!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png" width="895" height="597" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:597,&quot;width&quot;:895,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1325605,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://stories.logrocket.com/i/222513538?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!nIqf!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!nIqf!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!nIqf!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!nIqf!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4cf8cfe4-f3e2-4850-9105-35202037d962_895x597.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em>In our conversation, Hugh shares the principles he&#8217;s found to hold true across every large-scale transformation. He discusses why real transformation is always cross-functional and how the narrative around it needs constant advocacy. He offers insights on how to plan for turnover with an AI-powered knowledge base, the qualitative and quantitative signals that a multiyear program is on track, and how to tell when an external force like AI warrants a pivot. He also explains why the best transformation leaders aim to work themselves out of a job.</em></p><div><hr></div><h2>What holds true across every transformation?</h2><h3>Large-scale transformations can be driven by very different forces, including changes in the market, customer expectations, business models, technology, or company strategy. Across the transformations you&#8217;ve been part of, what principles have you found to be consistently true regardless of what&#8217;s driving the change?</h3><p>I think one of the axioms of transformation that rings most relevant is that they always need to be cross-functional. Usually, if you&#8217;re doing something very siloed, it&#8217;s not truly transformational. With that in mind, establishing the processes and working groups that extend not just beyond just your normal span of control but also your span of visibility is really important.</p><p>Ignoring those exogenous forces that you mentioned, the one universal truth I&#8217;ve found is that transformations are usually the result of something being ignored for a period of time. Maybe that&#8217;s tech debt or performance and scale needs, as you increase your customer base and move up the customer segment ladder toward large enterprise. Something that has been ignored will require deep investment into transforming your product, your infrastructure, maybe even your business model.</p><p>I like to think of it like a house. I live in a house, and I love the neighborhood. I like my neighbors. Maybe there are good schools, shops close by. I have a fit with this house, and this is analogous to product-market fit. Something clearly works, but as I grow and my family grows, our needs change, and that requires investment to meet our needs. If I want to continue to thrive, a renovation or a transformation would be needed.</p><p>One other truth that I&#8217;ve found is that transformations usually take place over a long period of time and can get political. The longer a project goes on, the more the opportunity exists for new, external inputs. When there are a lot of opinions or when a project gets political, there is a lot of opportunity for deviation from the original plan. So, it&#8217;s helpful to make sure at the outset you have incredibly clear goals and exit criteria &#8212; what the definition of transformed looks like. Without that, the goalposts can move an infinite number of times. And especially in very compliance-heavy areas, there are certain things that just must remain true.</p><p>Another thing that must absolutely be true at all times is that you have to be super sharp on your financials. Transformations are a capital-intensive investment that don&#8217;t often see the operational expense improvements for years potentially. Make sure you&#8217;re regularly, and very visibly, demonstrating adherence to the economics of your plan.</p><h3>Where do organizations tend to overcomplicate a transformation?</h3><p>That can happen when there&#8217;s a lack of clarity or a lack of alignment on what the North Stars or critical outcomes are. For example, how do you best demonstrate value on a transformation iteratively, and what does that value mean? Does it mean value into production right now? And, if so, across what metrics? Or does it mean the speed at which we&#8217;re able to deliver the end state of the transformation?</p><p>In this example, confusion or complexity arises because both outcomes are important, but you can&#8217;t satisfy both efficiently. So, you need to decide what&#8217;s the most important thing and advocate for that in a compelling way. You also need to factor in all of the different variables that might be going on in the background in terms of the size and stage of your company. Are you seeking funding? Are you publicly traded?</p><p>It goes back to having clear alignment of what the ideal end state is and the milestones along the way, and being really clear on the absolute basics. Between scope, time, and cost, you get two. What are the most important two at any given time? It&#8217;s OK if they change, but be realistic about the impacts of those changes, and the compounding effects of constant change.</p><h2>Owning the narrative</h2><h3>Getting people on board depends heavily on communication. How do you build a narrative that helps people understand not just what&#8217;s changing, but what it means for them?</h3><p>To avoid things getting contentious, constant, ongoing advocacy is really important. There&#8217;s the quote, &#8220;Nature abhors a vacuum.&#8221; If you don&#8217;t have a really strong narrative, someone else will fill that space with their own narrative, and then you&#8217;re rolling the dice. That&#8217;s when that complexity or confusion can arise.</p><p>When it comes to building a narrative that is compelling to users or to stakeholders, customers, whoever, the key thing there is not what you&#8217;re building, but why you&#8217;re building it. I tell my team that the messaging we give is the &#8220;so what?&#8221;, not the &#8220;what&#8221; &#8212; it&#8217;s about what it means to the user.</p><p>You build your business case for transformation all around metrics, but every metric has to be grounded in pain &#8212; pain for the business, pain for the user, pain related to compliance, whatever it may be. And that pain has an emotional resonance with your stakeholders. When you start with the pain rather than with what you&#8217;ve built, you communicate how it affects a human being, that lands so much more. That actually gets people&#8217;s attention because we are, by our own nature, averse to pain.</p><p>A really talented past leader of mine laid it out very clearly. Our organization is very compliance-heavy. We speak a lot of inside baseball, and we don&#8217;t realize that we&#8217;re doing it because it&#8217;s the language that we speak every day. You see this whenever you join a meeting, and there&#8217;s a lot of acronyms getting thrown around. They said flat out, &#8220;There&#8217;s too much inside baseball. You have to go with outside-in rather than inside-out communication.&#8221;</p><p>This goes back to how to build a compelling narrative: outside-in communication starts with the user, what they feel, and what matters to them.</p><h3>Given that transformation can have a long timeframe, how do you ensure messaging remains reinforced?</h3><p>Clear and compelling messaging is critical. Constant and ongoing advocacy is key for the success of your program. If you&#8217;ve got a year-long transformation, congratulations. You might have found a simple problem that requires a simple solution.</p><p>But with any multi-year transformations, the longer it goes on, the more effort you have to put into the internal product marketing, and the internal advocacy. You have to keep communicating why it matters and keep people glued to the original strategic outcomes &#8212; assuming they&#8217;re still relevant &#8212; in new, different, and novel ways. You&#8217;ll have executive attention for maybe three to six months. And then at the outset of the new year, you&#8217;ll be fighting for further investment in your program. So, you have to find new and different ways to continue to sell the idea of what you originally set out to do.</p><h3>You mentioned that longer transformations also tend to become more political. How do you prevent executive attention from waning?</h3><p>It&#8217;s a matter of reiterating and making it clear that the program, as originally set out, still meets the moment. And if it doesn&#8217;t, then you have to be very transparent about that as well. You should not get stuck in a sunk-cost fallacy.</p><p>A good example is the transformation that I&#8217;m working on right now. We started it just prior to 2025, before ChatGPT, Anthropic, Gemini became ubiquitous across the enterprise layer. Since then, our processes and strategy has had to pivot to account for that, and we were fortunate that the approach that we were taking ladders perfectly into agentic experiences and a future that is what we would consider AI native, not just AI adjacent.</p><p>But the narrative has had to evolve along with that to keep people excited about it and to build confidence that continued investment in this program will yield better outcomes for the business long-term than any other initiative.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive new posts every week.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h2>Planning for turnover</h2><h3>A transformation that lasts many years may not end with the team it started with. How do you keep the messaging consistent as new people join, and how do you make space for their new ideas?</h3><p>You will never have a shortage of new ideas and judgements on your old idea and decisions.</p><p>When you&#8217;re new at an organization, you have this sort of fresh-eyes approach on things, and you say, &#8220;Why do we do this this way? And why is that that way?&#8221; And then after you&#8217;re at a large organization for any period of time, you start to see how decisions are made, and the corners that you can get put into. You start to be a little less judgmental about past decisions and a little bit more solution-oriented toward the future. That being said, keep your fresh eyes perspective as long as you can after joining a new company, it&#8217;s valuable, and one of the reasons why you were hired.</p><p>But to your question on turnover, that&#8217;s unavoidable. If you&#8217;re at any organization working on a transformation program for two to four years, you&#8217;re going to experience turnover.</p><p>Wherever I go I like to create an internal knowledge base. Claude and ChatGPT agents have made this much easier. The knowledge base is something that we can put all of our documentation into. This repository links out to that information, so anyone at the company can ask any question. That&#8217;s been incredibly helpful, especially for ops, support, and product to avoid asymmetrical information. We operate in a highly complex, very deep space. The concentration of highly tenured domain experience in itself is a risk, not just to a transformation, but to a business. If information is only in people&#8217;s heads, you have to put it somewhere that is centralized. That&#8217;s the whole point of building something that is AI-native. It enables everyone, everyone becomes an expert.</p><p>We&#8217;ve found our knowledge base and other agents can aid in answering general inquiries and can also speed up onboarding. It can be a really helpful tool for giving newly onboarded employees with those really good ideas an understanding of the decisions and risks that have occurred on the program beforehand. That gives them context as to why an idea that they&#8217;re forming in their head with those fresh eyes may or may not be the right approach, without dissuading them from suggesting it.</p><p>One factor that&#8217;s really important in mitigating turnover during transformation is having strong executive-level champions. If you have any influence over who the sponsor of your program is, make sure it&#8217;s someone that really believes in not just what you&#8217;re doing but the way that you&#8217;re doing it. Ideally, they have the influence to keep the executives interested, aligned, in the same boat in terms of how the team is going to achieve this long-term transformation.</p><h2>Staying on course</h2><h3>What indicators, qualitative or quantitative, tell you a transformation is on track, especially when the end goal may be several years away?</h3><p>If you&#8217;re building iteratively and not in some sort of big-bang delivery methodology, it&#8217;s really helpful to have a leading indicator or lagging indicator. Like, how is user acceptance testing (UAT) going? We have our Customer Zero partners in the business performing UAT with us. So, ask: how are the vibes on UAT going? Is everybody feeling really good about the outcomes there? Is it matching their expectation of what our prototypes promised?</p><p>In terms of early quantitative indicators, I look toward delivery velocity. Are we at least on track from a delivery velocity perspective? Are we tracking toward the timeline?</p><p>Beyond that, are we balancing the right level of discovery and research in line with the delivery that needs to happen? For any long-term transformation program, you almost certainly have to take a dual-track agile approach. Delivery is important. We have to hit our timeline, but are we also able to chunk through the discovery, research, and early customer validation to turn that into deliverable requirements that are constantly validating that we&#8217;re building the right thing for our customers.</p><p>That is a little bit harder to measure. What does good discovery look like? What does good research look like? So, I would look to see how quickly we can put a prototype together and put it in front of the business or customers and say, &#8220;For this epic, for this feature or this experience, this is what we&#8217;re envisioning. These are the pain points that we want to hit.&#8221; How quickly can we get that validated to really move the needle faster on discovery and idea validation? Those are historically slower areas of the product development life cycle.</p><h3>You mentioned that the strategy of your current transformation had to pivot as AI tools spread across the enterprise. How do you know when an external force is significant enough to warrant a pivot?</h3><p>This is a completely different answer today than it would have been a few years ago. I think that anybody who started a long-horizon transformation program in &#8216;24 or &#8216;25 is likely pivoting to make sure that everything revolves around how AI incorporates into it, for better or worse. Not every company needs to consider AI, and not every company should, but every company is, just based on what I&#8217;ve seen in the market.</p><p>When is a pivot necessary? When the original outcomes or methods are no longer relevant to the moment. What sort of universal truths existed in the business plan and operating model, based around what metrics and what user pain? Do those still make sense?</p><p>Also, a pivot is necessary when leadership tells you that it&#8217;s necessary or when you tell leadership that it&#8217;s necessary. And I would bring this back to: If you let a gap in the narrative exist, someone will fill that for you. So, if you notice something within the market that materially impacts the approach that you would take or the outcomes of your transformation program, start crafting a new narrative, or at least start understanding how that impacts your outcomes. Are your metrics still relevant?</p><p>To use a hockey analogy: Are you skating to where the puck is going?</p><h2>Culture, scale, and the pace of change</h2><h3>How do corporate culture and company size shape a transformation?</h3><p>The two are inextricably linked. You can have a really strong culture at an enterprise, but there&#8217;s going to be dilution the larger the organization gets. If your culture is all about accountability and agency, well, at a 10,000-person company, that simply can&#8217;t exist everywhere due to hierarchies and approvals. But at a 100-person company, transformations can move much faster. They&#8217;re a lot easier because, philosophically it&#8217;s, usually it&#8217;s really easier to align people. There also tends to be faster decision making and a lot less churn.</p><p>If you have a corporate culture that&#8217;s oriented around status quo, or, &#8220;We&#8217;ve tried this in the past, and it didn&#8217;t work,&#8221; that&#8217;s going to make it materially more difficult to push a transformation through. This is why I come back to: Do you have the right executive sponsor with the right mindset leading the program? Because that can make all the difference. Even just one person in a sea of people who believes in what you&#8217;re doing, believes in the level of innovation that you&#8217;re trying to drive, believes in the outcomes that you&#8217;re trying to create, can help cut through the red tape and facilitate strong relationships with the executive.</p><h2>Working yourself out of a job</h2><h3>Is there one transformation that&#8217;s been the most rewarding for you?</h3><p>The most rewarding is the one I&#8217;m working on right now. This has been the most broad-scoped, intensive initiative, with lots of ideas floating around, and the most learnings that I&#8217;ve had on a program.</p><p>To give you a sense of the scale, our systems collect and distribute funds to the tune of hundreds of billions of dollars a year. We&#8217;re the leader in compliance within the HCM space. So, there&#8217;s a lot riding on the success of this transformation. There&#8217;s a lot of pressure on it, and rightfully so.</p><p>It&#8217;s fulfilling to observe how the teams swarm around the problems, and to see that we&#8217;re delivering against what we set out to do. But the most rewarding aspect of any transformation, not just this one, is the core architecture, platform, and the core capabilities that enable not just my team to operate better and faster but other teams to operate better and faster as well. Even though I don&#8217;t technically work on a platform team, I&#8217;m a huge advocate for core platform technology teams because of the critical role they play at scaling the entire organization&#8217;s ability to deliver great experiences.</p><p>At the end of your initiative, did you enable faster implementations for your customers? If so, that means that your revenue realization rates increased. Did you enable better large enterprise scale and performance, that unlocks sales to continue going after the absolute whales? Did you make any meaningful changes from a platform perspective that enable not just your team but the rest of the organization? Those are some ways that you can make a lasting change. I&#8217;ve always approached my work from the 100-year company philosophy. I want to make an impact that can be felt for generations, that&#8217;s why I get a lot out of transformation projects</p><h3>How do you sustain change after the transformation is complete?</h3><p>Building cross-functional working groups is an important part of a successful transformation, but so is making sure that embedded within those groups is decision authority. If you need an executive in every call to make a decision, you&#8217;re wasting time. If you want to avoid churn or spin, make sure that you&#8217;ve got decision authority in whatever meeting you are in, whatever ceremony you&#8217;re participating in, and make sure that those decisions are overcommunicated. Whether it&#8217;s a risk, a celebration, or a decision, always err on the side of overcommunication.</p><p>Good product leaders, product builders, and engineers who are focused on transformation are sort of like really good personal trainers. If you do your job well, you&#8217;re not needed anymore. I&#8217;m not trying to transform an organization so that I have a job there for 20 years. I&#8217;m trying to transform an organization, learn, help it be successful, and then go do something different, applying those principles and learnings in the same way. I want to work myself out of a job, and just go do the next thing after that even better. I think we should all approach transformation with that mindset.</p><h3>What does LogRocket do?</h3><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/?substack">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[“Give the Model a Layup”: How to Cut AI Cost Without Cutting Quality: LIVE Demo | Arjun Singh Jamwal (Verizon)]]></title><description><![CDATA[In a live demo of his app, Goodmuncher, Arjun explains why tokens are OpEx, how pre-processing lets smaller models perform like frontier ones, and offers a step-by-step demo for evaluating agents.]]></description><link>https://stories.logrocket.com/p/give-model-layup-how-to-cut-ai-cost-without-cutting-quality-arjun-singh-jamwal</link><guid isPermaLink="false">https://stories.logrocket.com/p/give-model-layup-how-to-cut-ai-cost-without-cutting-quality-arjun-singh-jamwal</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 27 Oct 2026 13:42:49 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/262960b9-9c17-4c91-8360-1ab65de2a7e4_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p></p><div id="youtube2-Ui3dGb0CmqU" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;Ui3dGb0CmqU&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/Ui3dGb0CmqU?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=Ui3dGb0CmqU">YouTube</a> | <a href="https://open.spotify.com/episode/4tsjwPgb1wbHRVZ02G29Sj">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/give-the-model-a-layup-how-to-cut-ai-cost/id1733103005?i=1000798353593">Apple</a></strong></em></p></div><p>Here&#8217;s something nobody warns you about when you build AI into your product: you pay for the work every single time a user touches it. If you&#8217;re a well-funded team, that&#8217;s just another line item. But for a bootstrapped founder, it can sink the business.</p><p>Arjun Singh Jamwal sees both sides of that. He&#8217;s a Senior PM at Verizon, where he&#8217;s worked on Verizon Cloud and now Verizon Home. On the side, he&#8217;s bootstrapping Goodmuncher, an app that helps people cut back on ultra-processed foods, and where every token he spends comes straight out of his own pocket.</p><p>His fix is something he calls &#8220;giving the model a layup.&#8221; Rather than having AI analyze every ingredient from scratch whenever someone scans a label, Goodmuncher scores ingredients and products ahead of time and keeps them in a vector database. Users get the same answer, and Arjun pays a fraction of what it would cost to run the analysis live.</p><p>This one&#8217;s a little different from our usual episodes: it&#8217;s a live demo. Arjun shares his screen and runs Goodmuncher&#8217;s content pipeline right there on the call. One command, 30 seconds, and out comes a finished 10-slide social carousel. Then he walks us through how it&#8217;s all built.</p><p>Along the way, we get into:</p><ul><li><p>Why you should think of tokens as OpEx, and how a little upfront prep keeps your AI costs predictable</p></li><li><p>How giving the model a &#8220;layup&#8221; lets cheaper models punch above their weight, and makes it easier to swap models later</p></li><li><p>Why he hard-codes the rules he can&#8217;t break instead of trusting the model to remember them</p></li><li><p>His step-by-step process for evaluating agents, from golden datasets to LLM-as-judge and observability tracing</p></li></ul><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. Why Arjun treats every AI token like an operating expense</h2><p>Plenty of companies are spending on AI simply because leadership told them to. Arjun doesn&#8217;t think that will last. Because those are real dollars, and sooner or later, that cost is going to add up.</p><p>Arjun thinks about it the way a finance person would: Every token you spend at runtime is a variable cost, and if you just throw raw data at a model and let it sort everything out, that cost gets hard to predict:</p><blockquote><p>&#8220;Think of tokens as a variable cost. This is like your OpEx. When you just throw everything at it to crunch the numbers, you&#8217;re actually working with an unpredictable OpEx.&#8221;</p></blockquote><p>The other option is to do some of the work up front, one time. Arjun thinks of that as CapEx.</p><p>And at some point, the math just stops making sense:</p><blockquote><p>&#8220;There&#8217;s no point in spending $200,000 on tokens when I could just hire a great full-stack engineer who could do that for me.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Treat AI costs like any other cost on your P&amp;L. If a feature uses tokens every time someone clicks, taps, or asks a question, your bill grows right alongside your user base. Before you launch, estimate what that feature would cost if usage grew 10x, so you aren&#8217;t caught off guard when the invoice comes.</p><div><hr></div><h2>2. How Arjun gives the model a layup by doing the hard work ahead of time</h2><p>Goodmuncher has to come up with a processing score for every food item a user looks up. The obvious approach is to have users snap a photo of the ingredient label, send it to a multimodal model, and let it work everything out from scratch. That&#8217;s fine with a handful of users. With 10,000, it gets expensive fast.</p><p>So, Arjun does the heavy lifting ahead of time. He pulls ingredient data from open sources, scores it in advance, and stores it as embeddings in a vector database.</p><blockquote><p>&#8220;What I&#8217;d rather do is pull in those ingredients, crunch that data, and already pre-attribute a processing score to it, so that when a user actually looks up the pesto sauce, for example, that score&#8217;s already ready.&#8221;</p></blockquote><p>Now the model isn&#8217;t reasoning through a long chain of steps. It&#8217;s just finding an answer that already exists.</p><p>Those savings add up. Arjun puts it at at least a 30X difference at scale, and says it&#8217;s let him cut what he plans to charge users by more than half.</p><p><strong>Product takeaway:</strong> Before you reach for a bigger, newer model, map out your workflow and ask which steps really have to happen in real time. Anything you can pre-process, cache, or look up ahead of time is work the model doesn&#8217;t have to do, and that&#8217;s money you can save yourself AND your customers.</p><div><hr></div><h2>3. Not every job needs a frontier model</h2><p>You don&#8217;t need the newest, most powerful model for everything. Frontier models shine on complex, multi-step reasoning. You can use them for simple tasks too, but you&#8217;re paying for horsepower you don&#8217;t need.</p><p>Keeping your workflows lean pays off in another way, too. When a better model comes out, you can swap it in without re-testing a long, tangled chain of steps:</p><blockquote><p>&#8220;When you have to swap out that model, the fewer steps you have in the workflow, the fewer steps of evaluations you need to run. So your evals will have a better time, your change management team will have a better time, your devs will have a better time.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Picking a model is a product decision, not just an engineering one, and the only way to know which one fits is to try a few. Get in the habit of testing the smaller, cheaper option first, and only move up when it can&#8217;t keep up.</p><div><hr></div><h2>4. How Arjun hard-codes the rules his AI should never break</h2><p>Goodmuncher is all about helping people eat less ultra-processed food, but it can&#8217;t claim to cure diseases or lower anyone&#8217;s medical bills. The FDA would have strong words for Arjun if it did. </p><p>During his live demo, Arjun shows how his Goodmuncher content pipeline checks every output against a &#8220;tripwire&#8221; file, a running list of banned phrases like &#8220;prevent dementia&#8221; or &#8220;save medical bills.&#8221; That&#8217;s not a call he leaves up to the model.</p><blockquote><p>&#8220;It&#8217;s important to make this deterministic in nature because we don&#8217;t want to compromise on this. These are the red lines that you cannot cross.&#8221;</p></blockquote><p>He also made a point of keeping that check separate from the step that writes the content, instead of baking it into the prompt:</p><blockquote><p>&#8220;I could have easily put my tripwire check into the content generation file, but I didn&#8217;t because I want it to be a different process, and I know this will build out over time.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Models follow the rules most of the time. But &#8220;most of the time&#8221; doesn&#8217;t cut it for compliance, safety, or anything that could hurt a user. Figure out what&#8217;s non-negotiable for your product and enforce it in your code, as its own step, so you can test it and add to it over time.</p><div><hr></div><h2>5. Arjun&#8217;s step-by-step process for evaluating agents</h2><p>Arjun mentors startups through the Cornell Tech community, and he keeps running into the same blind spot:</p><blockquote><p>&#8220;It&#8217;s so easy just to throw an agent at any problem. But what we&#8217;re not being mindful of is the eval aspect of it.&#8221;</p></blockquote><p>For Goodmuncher, this isn&#8217;t abstract. If a recipe recommendation ignores someone&#8217;s nut allergy, a real person could get hurt. Here&#8217;s how Arjun approaches evals:</p><ol><li><p><strong>Run an exploratory data analysis (EDA)</strong> on your production data to understand its categories and shape</p></li><li><p><strong>Build a golden dataset</strong> of 10&#8211;50 examples you know are 100% correct</p></li><li><p><strong>Reverse-engineer a rubric</strong> from that dataset (e.g., &#8220;Do the ingredients respect the user&#8217;s allergies?&#8221;)</p></li><li><p><strong>Create a staging database</strong> that represents production, with PII anonymized and de-identified</p></li><li><p><strong>Run evals against the rubric</strong>, using a mix of human evals and LLM-as-judge, with 5&#8211;10 examples per category</p></li><li><p><strong>Add observability tracing</strong> to pinpoint exactly where a workflow breaks</p></li><li><p><strong>Build a dashboard</strong> of those traces, if you can</p></li></ol><p><strong>Product takeaway:</strong> Building a golden dataset and rubric takes some work up front, but like everything else Arjun builds, that investment pays for itself every time you swap a model or chase down a bug.</p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=Ui3dGb0CmqU">00:00</a> Introduction<br><a href="https://www.youtube.com/watch?v=Ui3dGb0CmqU&amp;t=123s">02:03</a> Arjun&#8217;s product background<br><a href="https://www.youtube.com/watch?v=Ui3dGb0CmqU&amp;t=258s">04:18</a> Treating tokens as unpredictable OpEx vs. a One-time CapEx<br><a href="https://www.youtube.com/watch?v=Ui3dGb0CmqU&amp;t=514s">08:34</a> Pre-processing data so smaller (and cheaper) models can do the job<br><a href="https://www.youtube.com/watch?v=Ui3dGb0CmqU&amp;t=808s">13:28</a> Arjun's Demo: Turning one command into a 10-slide social carousel<br><a href="https://www.youtube.com/watch?v=Ui3dGb0CmqU&amp;t=1145s">19:05</a> Inside the pipeline architecture and its agents<br><a href="https://www.youtube.com/watch?v=Ui3dGb0CmqU&amp;t=1582s">26:22</a> Arjun's framework for evaluating AI agents<br><a href="https://www.youtube.com/watch?v=Ui3dGb0CmqU&amp;t=1927s">32:07</a> Conclusion</p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/arjunjamwal/">Arjun&#8217;s LinkedIn</a></p></li><li><p><a href="https://goodmuncher.com/">Goodmuncher</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at LogRocket.com.</p>]]></content:encoded></item><item><title><![CDATA[Your CEO Thinks AI is Working. Your Team Doesn’t. Here’s Why. | Annan Patel, VP Product (Tricentis)]]></title><description><![CDATA[Tricentis' VP of Product on why plug-and-play AI only gets you 60% of the way, why the remaining 40% is context nobody ever wrote down, and why the AI is never the one held accountable.]]></description><link>https://stories.logrocket.com/p/your-ceo-thinks-ai-working-your-team-doesnt-heres-why-annan-patel</link><guid isPermaLink="false">https://stories.logrocket.com/p/your-ceo-thinks-ai-working-your-team-doesnt-heres-why-annan-patel</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 20 Oct 2026 13:29:47 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/a9e1683c-d9ea-4a65-a601-401e5957867e_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p></p><div id="youtube2-JAYsxQ-SbFo" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;JAYsxQ-SbFo&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/JAYsxQ-SbFo?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo">YouTube</a> | <a href="https://open.spotify.com/episode/46289eRrxRkKD77BEQNqXf">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/your-ceo-thinks-ai-is-working-your-team-doesnt-heres/id1733103005?i=1000797253797">Apple</a></strong></em></p></div><p>Ask a CEO how their AI rollout is going, and 81% will tell you they&#8217;re highly confident. Ask the people doing the work, and that number drops to 56%.</p><p>Today&#8217;s guest sits right in the middle of that gap. Annan Patel is VP of Product at Tricentis, a leader in enterprise software testing, where he builds AI products for companies where one bad release can cost tens of millions of dollars. Before Tricentis, he spent more than 15 years in product across ad tech, Microsoft, and WP Engine.</p><p>Tricentis recently surveyed more than 2,500 people across the tech industry. 83% of companies said they trust AI agents to make release decisions, yet 60% admitted they ship untested code. The top reason? Pressure from leadership to move faster.</p><p>Annan&#8217;s view: AI gets sold as a tool you plug in. In practice, it&#8217;s a new operating system for how work gets done.</p><p>In this episode, Annan shares:</p><ul><li><p>Why plug-and-play AI only gets you about 60% of the way, and why the rest is context your team has never written down</p></li><li><p>Why Tricentis&#8217; first AI products stalled after a strong start, and how the team rebuilt them around one principle: the AI is never the one held accountable</p></li><li><p>Why tokens spent and lines of code shipped tell you nothing about whether your AI initiatives are adding value</p></li></ul><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. AI isn&#8217;t a new tool. It&#8217;s a new operating system.</h2><p>When AI companies pitch to executives, they speak the language executives already use: ROI, efficiency, cost reduction. This pitch is compelling, and the vision is real. But Annan thinks it misses what makes AI different from every SaaS product that came before it: it changes the <em>work</em> of the humans <em>doing</em> the work.</p><blockquote><p>&#8220;It&#8217;s not as straightforward as just adopting a new tool. In some ways, it&#8217;s adopting a new operating system. It&#8217;s a new way of working.&#8221;</p></blockquote><p>Teams that have worked together for years build up habits that don&#8217;t map neatly onto an agent. Annan says,</p><blockquote><p>&#8220;People develop ceremonies, cultures, practices in the way they work that don&#8217;t always easily translate to agentic or AI-driven systems.&#8221;</p></blockquote><p>That&#8217;s where the gap comes from. It&#8217;s not that anyone&#8217;s acting in bad faith. When you&#8217;re leading from 10,000 feet, you just don&#8217;t see the dozens of little fixes your team makes every day to keep things on track.</p><p><strong>Product takeaway:</strong> If your AI rollout plan looks a lot like a plan for buying new software, you&#8217;re probably underestimating it. You&#8217;re changing how people actually do their jobs, so give yourself time for the messy parts: rethinking processes, helping people adjust, and earning their trust. And when the folks closest to the work tell you something isn&#8217;t working, take them seriously. Because they spot the gaps long before you do.</p><div><hr></div><h2>2. The missing 40% is context nobody wrote down </h2><p>On paper, this sounds easy: write down your process, hand it to the AI, and let it do its thing. But the stuff that really matters on most teams was never written down in the first place.</p><p>When you work with a team for a long time, there are shorthands you develop. Annan saw this firsthand with Tricentis&#8217; own products. Customers said the AI was doing good work, but it only got them part of the way there:</p><blockquote><p>&#8220;What the AI is doing is good, but it maybe solves for 60% of what I need. Because that remaining 40%&#8230;comes from information about my business, my team, my industry, the way that we work, the history we&#8217;ve built up over time.&#8221;</p></blockquote><p>You can&#8217;t fix that by pointing an agent at a repo or a SharePoint folder. That context lives in people&#8217;s heads, so the system has to learn it from them.</p><p><strong>Product takeaway:</strong> Before you scale an AI workflow, sit down with your team and get the unwritten stuff out of their heads. This includes the edge cases, the incidents everyone still remembers, the compliance rules, and all the other habits. If your AI output keeps coming back generic or almost right, it&#8217;s probably missing context, not a bad model.</p><div><hr></div><h2>3. Start with one job to be done, and treat it like CapEx</h2><p>Annan has noticed a pretty clear pattern in which companies make AI work and which ones end up frustrated.</p><blockquote><p>&#8220;Where organizations get a bit frustrated is when they try to do too much at once&#8230; Because that&#8217;s when you start to identify the skeletons in the closet.&#8221;</p></blockquote><p>What works better is starting small. Pick one job, figure out how an agent can help speed it up, and then put the time you get back toward something else. Tricentis did this internally with security fixes, which let the team move fast in one area without blowing up its whole development process.</p><p>The catch? Those early wins usually won&#8217;t match the big ROI numbers executives hear in sales pitches. You only get those by changing your core business processes, and that takes more money, more time, and a lot more care.</p><blockquote><p>&#8220;Sometimes you have to slow down to move fast.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Keep your first AI projects small enough that they can work without overhauling everything. For the bigger transformation, think of it as a long-term investment. You probably won&#8217;t see the full payoff in the first year, but you&#8217;re laying groundwork that pays off later. Just make sure leadership knows that going in.</p><div><hr></div><h2>4. Tokens and lines of code are the wrong scoreboard</h2><p>When teams felt pressure to prove their AI spend was worth it, a lot of them reached for whatever numbers were easiest to count.</p><blockquote><p>&#8220;We&#8217;re not always measuring the right thing&#8230; What you saw a lot of a year ago was the tokenmaxing, the &#8216;Hey, how many lines of code am I shipping?&#8217;&#8221;</p></blockquote><p>The problem shows up later, in code that gets rolled back, ships without being tested, or never makes it to production at all. Testing teams are feeling it most. They have fewer people and, in some cases, 10x more code to check.</p><p>And as Annan points out, leadership doesn&#8217;t really care about those numbers anyway:</p><blockquote><p>&#8220;At the end of the day, the executives don&#8217;t care how much AI was used, they care, &#8216;Did I get to the outcome that I needed?&#8217;&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Counting tokens or lines of code tells you how busy everyone is, not whether anything got better. Measure AI against the things your business already cares about, like quality, revenue, and whether customers actually use what you ship. As Annan puts it, building faster makes good product judgment matter more, not less, because <strong>you can build the wrong thing really fast.</strong></p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo">00:00</a> Introduction<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo&amp;t=137s">02:17</a> Annan's path from ad tech to enterprise testing at Tricentis<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo&amp;t=224s">03:44</a> Why executives and teams see AI's capabilities differently<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo&amp;t=599s">09:59</a> From coding assistants to agents to fully unattended workflows<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo&amp;t=787s">13:07</a> AI hype on LinkedIn vs. the real risks enterprises face<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo&amp;t=1090s">18:10</a> Why tokens and lines of code are the wrong ROI metrics<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo&amp;t=1199s">19:59</a> How Tricentis adopts AI internally, from security fixes to PM work<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo&amp;t=1543s">25:43</a> What "think like a CEO" really means for product leaders<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo&amp;t=1816s">30:16</a> When velocity matters and when quality wins, depending on your market<br><a href="https://www.youtube.com/watch?v=JAYsxQ-SbFo&amp;t=1967s">32:47</a> Conclusion</p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/annanpatel/">Annan&#8217;s LinkedIn</a></p></li><li><p><a href="https://www.tricentis.com/">Tricentis</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at LogRocket.com.</p><p></p>]]></content:encoded></item><item><title><![CDATA[Leader Spotlight: Rethinking PLG for probabilistic products, with Lauren Schuman]]></title><description><![CDATA[Lauren Schuman is VP of Product Growth at Close, a sales CRM for small scaling businesses, where she is building out the new product-led growth (PLG) function and working on setup, activation, and monetization for new AI-native features.]]></description><link>https://stories.logrocket.com/p/leader-spotlight-lauren-schuman</link><guid isPermaLink="false">https://stories.logrocket.com/p/leader-spotlight-lauren-schuman</guid><dc:creator><![CDATA[Jessica Srinivas]]></dc:creator><pubDate>Thu, 15 Oct 2026 07:02:50 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!1M4i!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><em>Lauren Schuman is VP of Product Growth at Close, a sales CRM for small scaling businesses, where she is building out the new product-led growth (PLG) function and working on setup, activation, and monetization for new AI-native features. She has over 20 years of experience in product and growth. Most recently, she was VP of Product Growth at Bitly, where she managed a self-serve P&amp;L of over $75M in ARR, and before that at Mural. As Senior Director of Product Insights &amp; Growth at Mailchimp, she started the company&#8217;s Growth discipline and led the Product Insights and Growth teams, contributing to revenue growth from $500M to over $800M. She began her career as a buyer in corporate retail.</em></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!1M4i!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!1M4i!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!1M4i!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!1M4i!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!1M4i!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!1M4i!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png" width="895" height="597" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:597,&quot;width&quot;:895,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1333016,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://stories.logrocket.com/i/219780808?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!1M4i!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!1M4i!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!1M4i!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!1M4i!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F152d941d-9e6e-4a6c-a3fc-d61a761328ec_895x597.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em>PLG was built around predictable software. What happens when the product itself becomes dynamic, personalized, and probabilistic? When Lauren joined LogRocket&#8217;s LaunchPod podcast in 2025, she predicted that traditional onboarding playbooks would give way to natural-language, genuinely personalized experiences, and that faster AI-driven learning would create new bottlenecks, particularly around discovery and adoption. She predicted it, and now she&#8217;s building it.</em></p><p><em>In our conversation, Lauren discusses why AI can pull teams away from growth fundamentals, how a widening tech-savviness divide is reshaping onboarding and adoption benchmarks, and what it takes to earn trust in products that won&#8217;t be right all the time.</em></p><div><hr></div><h2>Leading with the problem, not the technology</h2><h3>You&#8217;ve built PLG engines around established concepts like activation, time to value, conversion, and retention. How does that change when you apply them to AI-enabled features?</h3><p>What I&#8217;ve been finding is that some of the new technology is distracting us away from the core principles. You&#8217;re so eager to say, &#8220;What could an agentic onboarding experience look like?&#8221; that you&#8217;re sort of forgetting the whole point of what you&#8217;re trying to solve by leading with the technology.</p><p>With my team right now, we&#8217;ve talked a lot about understanding what is possible, but not jumping to what technology to use to solve the problem. I give credit to Max, our senior engineering manager, who&#8217;s done a lot of pushing to think about the problem we&#8217;re trying to solve and not the technology, and have the engineering team partner with us on the best way to think about implementing the solution.</p><p>A lot of the fundamentals are still important. What are the jobs that the customers are trying to accomplish? Those jobs pretty much haven&#8217;t changed in nature. How they accomplish them may be changing. But at the root of it, if you still think about what a prospect or customer is trying to accomplish, how they&#8217;re feeling in that moment, and the steps that it takes, I think you ultimately end back at a similar place.</p><p>You actually have to be a lot more strategic. We&#8217;re saying, &#8220;What is the right vehicle to solve a problem at a specific point in time, and what is the right approach?&#8221; There are more and different vehicles and approaches to consider, but staying rooted in the fundamentals &#8212; what is the customer trying to accomplish, what are the things that make them successful in your product &#8212; is really important.</p><h2>A widening divide in tech-savviness</h2><h3>Where are you seeing things actually change?</h3><p>I don&#8217;t have the evidence to necessarily support this yet, but I feel like there&#8217;s a larger divide in tech capabilities, especially for small businesses. You&#8217;ve always had a continuum of early adopters and then your laggards. I think now, with the speed of innovation and the accessibility of certain technology, that gap is widening. That makes it even more difficult to figure out: Should your product be well-suited for the early adopters, should it be suited toward the laggards, or are most people in the middle? Those experiences might look entirely different.</p><p>If we&#8217;re talking about time to value or activation for someone who&#8217;s super tech-savvy, they might be able to get up and running having a conversation with an AI agent, and that&#8217;s a no-brainer. But somebody who really doesn&#8217;t understand how agentic technology is working, or any of those patterns, or how to do good prompting &#8212; that&#8217;s an entirely different experience that might be needed to serve that type of customer. What I&#8217;m getting at is that I think you have to be even more segmented in your approach. It&#8217;s not a one-size-fits-all.</p><p>In some ways, the growth practice was built for this, because testing and learning quickly has always been important, but now it&#8217;s even more important. If you go listen to any of the companies that are thriving right now, they&#8217;re all like, &#8220;Our speed of learning is just so fast. The faster that we can learn, the better that we&#8217;re doing.&#8221; Companies that are not having that rapid speed of testing and learning toward what&#8217;s really new are the ones I think are sort of falling behind.</p><p>One other thing I would add is that trust is even more important. With deterministic products, we knew we could control the outcome: do this and, most of the time, you get that. But now that products are more often not deterministic, we&#8217;re like, &#8220;Ugh, how do I get you to trust this thing that may not be correct all the time?&#8221;</p><p>And by the way, what is acceptable? You&#8217;re actually having to teach customers how to use the product that isn&#8217;t going to be right all the time, and what to accept, and what is good and what is not good. The core principles remain the same, but the tools that you have are a bit different, and there&#8217;s probably even more to consider around how comfortable people are with technology and teaching them alongside you.</p><h3>When different segments use the same feature in different ways, how do you measure success?</h3><p>I think it&#8217;s more about being honest about your benchmark for adoption. We launched a voice AI agent, and while we worked at making it as approachable as possible, we realized there are some people who will be entirely intimidated by the idea of answering any questions to set up a voice AI agent.</p><p>Some of this factored into our initial thought of what success would look like. Success wasn&#8217;t going to be a mass market adoption. Success was going to be like, &#8220;Hey, we had X amount of customers that were really using this product and getting a ton of value out of it. They were raving fans, and they were telling everybody about it.&#8221; Maybe that was only 5 or 10% of our core customer base, versus another feature that is much more widely distributed, and you would expect 50% of your customers to use it. Some of it is in the features that you&#8217;re building up front, and who they&#8217;re aimed at in terms of that tech-savviness.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive new posts every week.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h2>Onboarding by intent</h2><h3>Given these different segments, do you think about onboarding differently? Do you set up different flows for different types of customers?</h3><p>We are very actively working on this, and we have been thinking about the different channels and modalities in which people would want to learn. We&#8217;ve also been thinking a lot about what it means to get started with a product for different people.</p><p>For you and me, we might have very similar businesses, but you want to come in and just kick the tires. You want to click around, you want to test things, you want to make sure the features exist. You&#8217;re OK with maybe using fake data, or something that just gets you a sense and a feel, and that&#8217;s how you make your purchasing decision.</p><p>I might be more of the camp of like, no, I actually need to try it out. I need to have real customers in, or real prospects in, added as leads. I need to make a real phone call, see how the phone system works. I need to send an email. I actually need to see it in action.</p><p>We&#8217;ve been talking a lot about this related to intent. Are you there to explore? Are you there to actually set up and get going with the product? What is the best way to meet the needs of somebody who&#8217;s exploring versus somebody who is ready to set up and get going? If I&#8217;m exploring, maybe I need a lot more different paths that I can possibly take, because my chosen path might depend on what features matter to me. Maybe a conversational approach would be great, because it&#8217;s more open-ended in nature.</p><p>We tried using our AI chat agent, and we created a version of it that was like an onboarding coach. We said, &#8220;What would happen if the onboarding coach was the first thing you saw, and we had a couple of prompts that were like, &#8216;Tell me how to do these particular things,&#8217; right off the bat? How would that work?&#8221; I think we had roughly 20% of customers engaged with the chat. But the experience was just like, &#8220;Here&#8217;s some information, go figure out what to do.&#8221; It wasn&#8217;t at the point where we were doing actions on the customer&#8217;s behalf yet, so I don&#8217;t know that it was super valuable.</p><p>On the flip side, you might have folks who are ready to come in and immediately set up. What are things that we can do to support you in that process? Should we be super opinionated on what the first step is for a customer like you? Does that mean we actually create workflows on your behalf &#8212; you give us a few pieces of information, we go ahead and set it up, and you get going? What level of control is needed?</p><p>I&#8217;d love to say I&#8217;ve solved this problem. I haven&#8217;t solved it, but I&#8217;m thinking a lot about the idea of intent mixed with level of control versus opinion, and then what&#8217;s possible to either guide you or to do on your behalf.</p><p>Some of the principles we&#8217;ve been working with are: &#8220;tell me,&#8221; then there&#8217;s &#8220;do it with me,&#8221; and then &#8220;do it for me.&#8221; How can we pay off each one of those? I don&#8217;t imagine that everybody will want to fall in all three camps at all different times, but how do we start to feel our way into the areas where people do want it to do it for them versus do it with them?</p><h3>What signals do you use to decide which flow to put someone in?</h3><p>Right now, we are actually going to ask people their intent. We&#8217;ve tried to infer some of that, but when I originally came in and took over, we had a very long signup flow. It was like 12 steps to get to the product. We cleaned that up and basically said, &#8220;OK, you have to earn the right to ask questions.&#8221; We&#8217;ve got it down to two steps into the product. We&#8217;re adding one or two back. Once you get into the product, we&#8217;re adding back a question about your intent, and then one about which channels you&#8217;re interested in using, and we&#8217;re using that to personalize the experience.</p><p>We&#8217;re actually about to launch an experiment around this where, based on your inputs, you literally watch your onboarding recommendations populate in real time, so you can see that payoff for you. And then we&#8217;re doing a whole simulation for phone number setup. Say you are here to explore and you do make phone calls. We know that phone calling is one of the most valuable parts of our product. If we could just get people to experience what it&#8217;s like to make a phone call in the product, then they will be much more likely to get started. We&#8217;re setting it up where you get a real phone number, you make a phone call, and your very first lead is actually our AI assistant. You call our AI assistant, you have a conversation, you watch the CRM update with all the details that shows you exactly what would happen, and then you can do it for yourself with your own information.</p><p>Those are the kinds of things that we&#8217;re exploring: How much can you infer versus when is it useful to actually ask questions and then show the personalization payoff in real time?</p><h2>Trust in a probabilistic product</h2><h3>You mentioned trust earlier. What practices do you rely on now to build trust, or to avoid eroding it?</h3><p>That&#8217;s super hard. I think trust is more important than ever before. In the past, trust was logos of companies that you do business with, social proof, those types of things. That&#8217;s still relevant, but what&#8217;s really difficult about the trust piece is that the bar for &#8220;this is no good&#8221; is different for everybody.</p><p>Say you&#8217;re interacting with an AI chat product and you have one bad experience. That could be enough for you to say, &#8220;I no longer trust this.&#8221; But it&#8217;s really unpredictable in terms of what that will be across the different customer base and whether or not you&#8217;ll have it.</p><p>I know a lot of product teams are working on evals to say, &#8220;How often is this getting these things right, and what is the quality?&#8221; But it&#8217;s still to the point where your product could be good 95% of the time, but for that 5%, one thing happens and all of a sudden somebody doesn&#8217;t trust this product or feature anymore. How much that extends to the rest of your brand and your product is not really super well known.</p><p>I actually think it&#8217;s really important to use AI where it is the best solution, but not all the time. Sometimes a programmatic automation is actually going to be a better choice than having an AI agent do it, because it is predictable. Maybe it&#8217;s like 90% good, but maybe that predictability is actually better than the real wow factor and the risk that&#8217;s associated with the agentic version.</p><p>So I think it&#8217;s about being really smart about where you apply AI, and helping benchmark. If you can provide benchmarks for customers, that lets them know how good something is. It&#8217;s also about getting folks in the mindset that they can help improve your products. If you could get somebody invested enough to care in providing feedback to make your product better, and then seeing improvements, that can help build trust.</p><p>There is such a thing as speed to trust, or initial trust, and that&#8217;s one benchmark. But then there&#8217;s the ongoing risk that you could earn trust, and then something happens, and all of a sudden the trust for that product is gone. I don&#8217;t have a good answer for how to measure that, and also how to get people to come back and give your product another try.</p><h2>Pricing in the wild, wild west</h2><h3>How are you thinking about pricing and packaging for AI features?</h3><p>We keep saying pricing is like a wild, wild west, and I think nobody&#8217;s figured it out. The thing that I do like about it is that we&#8217;ve all agreed that you have to test and learn regularly. Back when I joined Mailchimp they hadn&#8217;t touched their pricing and packaging in 10 years. Nowadays, can you imagine? The best companies are rapidly iterating regularly, and I think that&#8217;s new and different.</p><p>I started my career in corporate retail, so I was a buyer in the apparel world. In that world, everything is about cost plus margin, so you&#8217;re very used to having a cost of goods sold. As I moved into SaaS, it was like people just put your finger up in the air and be like, &#8220;I think this feature&#8217;s worth this amount and I&#8217;ll price in this way,&#8221; because you had no real costs. It was really your people costs. Now I think the industry is learning what it&#8217;s like to actually have costs associated with technology.</p><p>The mistakes I&#8217;m seeing are where it&#8217;s just like, I&#8217;m adding a markup to whatever the cost is and passing it along to the customer. I think that doesn&#8217;t work well, because that may not equate to value. Just like in a grocery store or a retail store, you&#8217;re going to have things that maybe lose money, but they&#8217;re table stakes and you have to carry them, and there&#8217;s a willingness to pay that maybe doesn&#8217;t match the cost. There are other things where maybe the cost is actually quite low, but the willingness to pay or the value is super high.</p><p>Yes, you need to think about cost, but you should make decisions according to the value, not necessarily just the cost. I&#8217;ve noticed that&#8217;s often a really hard muscle to build for folks who don&#8217;t come from the world of actually thinking, &#8220;It&#8217;s not just about, &#8216;This thing costs this much, so therefore we&#8217;ll pass it on with a 20% markup or whatever.&#8217;&#8221; You have to consider how it all comes together.</p><h3>How are you approaching monetization for your own AI features?</h3><p>Right now, our AI features are included in our plans, so you don&#8217;t have to buy them separately. You get AI credits as part of your plans, but if you are using above a certain threshold, that&#8217;s when you would purchase an AI subscription or eventually a top-up package.</p><p>We purposely did that because we didn&#8217;t want monetization to get in the way of adoption. I think right now, our biggest hurdle is just more awareness and adoption of the AI features, and we&#8217;ve given a lot of leeway for folks to try our features. Our goal upfront was not to try to get monetization off of it. It was more about, &#8220;Hey, is this feature actually useful to you?&#8221; If it&#8217;s useful to you, it&#8217;s going to be a no-brainer that you&#8217;ll want to buy more credits, because you&#8217;re clearly getting a lot of value from it.</p><h2>Protecting the core</h2><h3>Are there assumptions you had about building with AI that have proven untrue?</h3><p>There&#8217;s so much hype about what the technology can do that I thought everybody was going to run super fast and, for instance, have agentic onboarding experiences. I thought that classic product as we know it, where you have a UI and you&#8217;re clicking UX elements and doing your tasks that way, would be a pattern of the past.</p><p>But I don&#8217;t think most products have gone there. Many people have added in AI agents and agentic capabilities, but I haven&#8217;t seen the uber personalization payoff yet in many of the products. I&#8217;ve seen a lot of add-on AI agents. Many of them are not good. I still think a lot of the most advanced users go to Claude, and they hook up via MCP, and they&#8217;re using products through Claude as the main interface. So maybe that&#8217;s the learning: How much do you build out in your own product versus how much do you open your product to be consumed by other products?</p><p>I think everybody&#8217;s grappling with this: If it&#8217;s open, then what&#8217;s the true value prop in your product if any of a number of products could have the same access and create the same thing? I think the best teams are thinking carefully about the best jobs to solve with AI, not just trying to use AI to solve all the jobs.</p><h3>How should teams balance building new AI experiences with the product most customers still use?</h3><p>There is the notion of &#8220;be where people already are&#8221; versus trying to fight that. If there&#8217;s a subset of your base that&#8217;s already in Claude, and that&#8217;s how they want to run their workflow in their day-to-day, don&#8217;t fight against that. What would be the right strategy to monetize that, or what&#8217;s the unique value there? But that segment is not going to be 100% of your customer base, so how do you evolve who you&#8217;re looking for and servicing?</p><p>The big thing that I&#8217;m wondering with a lot of companies is, if we&#8217;re all so invested in building new AI features or being AI native, what about the rest of your product experience? What&#8217;s happening to it, and are people paying attention to it? I can&#8217;t overstate the number of companies that I&#8217;ve been in where they&#8217;re so focused on innovation and the next thing and the next thing and the next thing that they forget their core product and their core customers. Then that becomes stale, or broken, or out of date, and they can really lose foothold because they&#8217;ve lost sight of that.</p><p>I&#8217;m not saying you should bury your head in the sand and not innovate or do anything new and interesting. But at the same time, I wonder what percentage of organizations&#8217; teams are focused on all the new versus still making sure the core does what it needs to do. I bet if you were to ask a lot of product leaders right now about the percentage breakdown of who&#8217;s using their core product versus their AI features, the answer would still be overwhelmingly their core product.</p><h3>What does LogRocket do?</h3><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/?substack">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[Leader Spotlight: The player-coach approach to product leadership, with Uday Marepalli]]></title><description><![CDATA[Uday Marepalli is Director of Product Management at Upwork, a $4B marketplace, where he leads client outcomes and retention.]]></description><link>https://stories.logrocket.com/p/leader-spotlight-uday-marepalli</link><guid isPermaLink="false">https://stories.logrocket.com/p/leader-spotlight-uday-marepalli</guid><dc:creator><![CDATA[Katie Schickel]]></dc:creator><pubDate>Wed, 14 Oct 2026 07:02:44 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!iamH!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><em>Uday Marepalli is Director of Product Management at Upwork, a $4B marketplace, where he leads client outcomes and retention. He has spent the past 10+ years across AI, marketplaces, consumer, and SaaS, beginning his career in technology venture capital. Uday later worked in the CEO&#8217;s Office at Blaize, an AI computing platform. Before joining Upwork in 2021, he led jobseeker personalization for 50M+ users at Glassdoor.</em></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!iamH!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!iamH!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!iamH!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!iamH!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!iamH!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!iamH!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png" width="895" height="597" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:597,&quot;width&quot;:895,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1297107,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://stories.logrocket.com/i/219781143?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!iamH!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!iamH!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!iamH!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!iamH!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3debc6ef-e262-4ce8-ac6b-e75bd3193c09_895x597.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em>In our conversation, Uday shares his player-coach philosophy: how product leaders can stay close to customer problems while scaling their impact across teams. He discusses balancing hands-on leadership with autonomy, coaching PMs to develop judgment and taste in an era of AI, and maintaining strategic focus as product development accelerates.</em></p><div><hr></div><h2>The player-coach philosophy</h2><h3>What is your philosophy on a product leader&#8217;s role as player-coach?</h3><p>I directly own a few things end to end and lead everything else through other people. A product leader&#8217;s job is to raise the ceiling on the team&#8217;s thinking, and you cannot do that entirely from a distance. I retain some direct ownership to stay calibrated. Product management depends heavily on judgment and taste; if you stop doing the work, your feedback becomes less valuable. People may still act on it because of positional authority, which can be dangerous.</p><p>I tend to step in directly when an important problem lacks a clear owner, such as an outdated experience, a gap between teams, or a zero-to-one problem that is not yet well enough shaped to hand off. I coach when the goal and ownership are clear, but the PM needs a thought partner, sharper judgment, or help removing obstacles.</p><p>A leader who only coaches eventually loses touch with the craft, while a leader who only plays is limited by their own capacity. Product leaders today need to deliberately balance both roles.</p><h3>As a PM moves into leadership, how do you suggest they stay close to customer problems?</h3><p>Product always starts with the customer. As you move into leadership, staying close to customers must become intentional, because it no longer happens automatically. Your calendar fills with internal alignment, reviews, planning, and recurring meetings. If you do not protect time for customers, you can go months without hearing from one directly.</p><p>Secondhand research is useful for identifying patterns, but it compresses the experience. You receive a conclusion, without experiencing the hesitation, frustration, or intensity behind the customer&#8217;s pain. A slide might say customers find something confusing, but hearing someone struggle with the experience for five minutes gives you a very different understanding of the problem. I maintain a regular cadence of customer conversations and talk to a customer every couple of weeks.</p><p>I use the product I build and stay up to date on the end-to-end flows, even when my teams do not own them. I do not need to join every interview, but talking to customers and using the product firsthand can help me calibrate the insights that reach me.</p><p>That proximity to customers also makes me a better coach. When I understand what customers need, I can help a PM distinguish between a feature request and the underlying problem. The goal is not to bypass the team&#8217;s research, but to maintain an independent view of the customer so my judgment stays grounded in reality.</p><h3>Can you give an example of what the player-coach approach looks like in practice?</h3><p>One example was the Early Access Customer Program we launched at Upwork earlier this year. Many teams were shipping new experiences, including a new design system and agentic products, to small audiences so they could gather customer feedback and iterate quickly.</p><p>Each team independently recruited customers from our broader user base. Every team was doing its job, but the system wasn&#8217;t working well. Recruitment was inefficient, customers had no clear way to understand or manage their participation in experimental products and features, and teams lacked a scalable pool of engaged testers. It was a cross-team problem that no single PM was positioned to own.</p><p>I stepped in to shape the solution. I wrote the initial requirements and worked with design, product marketing, launch, and product operations. Together, we created the centralized Early Access Program, with clear ways for customers to opt in to and out of early feature testing.</p><p>Just as important, I handed the program off after launch. Product operations took over, multiple launches have since used it, and it has become a repeatable launch model for the company.</p><p>That is the player-coach model: step in when a cross-team problem needs leadership, create momentum, and establish durable ownership.</p><h2>Hands-on leadership without micromanaging</h2><h3>Where&#8217;s the line between being hands-on and micromanaging?</h3><p>It is a fine line. For me, the distinction comes down to which aspects of the work you examine closely and why. I focus on the customer problem, the supporting evidence, and how the work connects to company goals. Leadership context and judgment can materially improve the product&#8217;s direction in these areas.</p><p>Once we align on those points, I give the PM room to shape the hypothesis, solution, and execution. Being hands-on means adding signal that improves the outcome. Micromanaging means leading with authority or personal preferences in ways that do not benefit the team.</p><p>For example, if I say, &#8220;Can we change this copy because I prefer another phrase?&#8221; or prescribe every step without evidence, I am probably not helping the team. But if I ask, &#8220;What problem are we trying to solve? Why now? How will we measure success?&#8221; I can help the team sharpen its goals and execute more effectively.</p><p>One practical habit is to state explicitly which mode I am in. Sometimes I say, &#8220;This is one person&#8217;s opinion. Feel free to discard it.&#8221; Other times, I say, &#8220;I have a strong recommendation.</p><p>Let&#8217;s debate it, and one of us should leave convinced.&#8221; Occasionally, I make a mandate involving strategy, risk, or a decision for which I am directly accountable. In those cases, I am clear that I am making the decision and will own the outcome if it does not work.</p><p>When a problem lacks clear ownership, a player-coach should step in, create a system, hand it over, and move on. You cannot become attached to something you created because you must preserve the capacity to address the next important problem.</p><h3>How does staying close to product details make you a better coach to experienced PMs?</h3><p>There is a common leadership phrase: &#8220;Hire great people and get out of their way.&#8221; I agree that hiring great people is critical, but the second half is often oversimplified.</p><p>Experienced PMs generally do not need someone to teach them the basics of building a product, but they value a credible thought partner. You cannot play that role if you are not close to the details, because your understanding of the product will be superficial.</p><p>Staying close to the details helps me ask sharper questions and offer more relevant context. If I notice something broken and raise it thoughtfully, it signals that I care about the work and understand the team&#8217;s constraints. That builds trust. Spotting an issue does not mean prescribing the answer.</p><p>I also create low-stakes spaces where PMs can riff on unfinished ideas without feeling evaluated. The goal is to help them sharpen their thinking and ultimately ship a stronger product for customers.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive new posts every week.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h2>Coaching through judgment and AI</h2><h3>You mention judgment as a critical skill with AI. Are you finding you have to push back on PMs more now, as they rely heavily on AI to produce their output?</h3><p>AI has made producing work much cheaper, which makes judgment more valuable. Writing a document, analyzing data, and building a prototype used to be significant bottlenecks; now they are much faster to produce. The scarcer skills are deciding what deserves to exist and whether the output is actually good.</p><p>That said, AI produces many poor first drafts. It can generate polished, confident work that is still wrong. As output has increased, my coaching has shifted toward verification and testing. If a PM says, &#8220;I created this analysis using AI,&#8221; that is great. People should use AI to work more efficiently. My coaching then turns to questions: How do we know this is true? What would disprove it? Is it merely well presented, or is it genuinely good? I act as an editor, helping people reach a polished artifact we trust.</p><p>This is where being hands-on helps. If you understand how users move through a flow, you can flag engagement data that does not make sense. If you know the customer pain points well, and the synthesis seems counterintuitive, you can review the source transcripts to understand what customers actually said.</p><h3>How are you using AI to shorten the customer discovery loop without losing firsthand understanding that comes from talking to customers?</h3><p>I use AI to accelerate the work around and after customer conversations, but I prefer to conduct the conversations myself. Before an interview, I use AI to review account details, draft a discussion guide, sharpen my hypothesis, generate counter-hypotheses, and stress-test my thinking.</p><p>As a human, I might be attached to an idea, but AI is not. I can use it to broaden the range of possibilities or ask plainly why my hypothesis might fail. Then I enter the customer call ready to probe the real problem instead of orienting myself or projecting my own assumptions.</p><p>AI also makes it faster to create something concrete, such as a prototype for customers to react to. We recently explored a marketplace buyer protection program, and I moved from an initial hypothesis to customer insights in under a week. A few years ago, that would have taken much longer.</p><p>After the interview, AI can help synthesize findings and identify patterns across conversations. I use it carefully because it can flatten the intensity of customer feedback. Before packaging or acting on recommendations, I review the original transcripts myself for a final check.</p><h2>Scaling product leadership across teams</h2><h3>As your scope grows, how can you use your view across teams to identify and solve problems that individual PMs may not see?</h3><p>The main advantage of broader scope is stronger pattern recognition. You see where customer journeys break across team boundaries and where sensible local decisions create a poor system-level outcome. Individual PMs are rightly focused on making their areas successful. Product leaders must connect those efforts into a coherent customer journey. The seams between teams often hold the largest opportunities.</p><p>I stay close by reading each team&#8217;s updates, reviewing metrics across the portfolio, and looking for recurring customer behaviors or tactics through firsthand product use and customer conversations. I currently lead our client-retention charter. Some teams help active customers do more, while another focuses on reactivating customers who have left the platform.</p><p>These teams target different customers, but their tactics, customer needs, and platform capabilities overlap. A cross-team view allows me to identify opportunities for them to partner.</p><p>My role is to connect those intersections, align teams around the shared problem, and establish clear ownership so each team has its own swimlane.</p><h3>What mental model do you use to think about being a player-coach at a time when AI is increasing individual output?</h3><p>Production has become much cheaper, making judgment and taste more valuable. Historically, the main objection to hands-on leadership was capacity: two days spent building something were two days not spent leading the broader organization. With AI, the same task may take an afternoon, so much of that trade-off disappears. I can analyze a funnel, build a prototype, or conduct a competitive teardown without taking ownership away from the team.</p><p>I can create a rough artifact and ask my team what is missing. That gives us a more productive starting point while preserving their ownership. It also makes the editor&#8217;s role more important: countering AI slop and coaching for verification and taste.</p><h2>Building judgment and choosing the right problems</h2><h3>How are people coming up now as PMs who didn&#8217;t have experience working without the assistance of AI building taste and judgment?</h3><p>This is where mentorship matters. I have found that the best PMs are always open to feedback.</p><p>Product management is largely an apprenticeship: you learn what good looks like through experience and from mentors you can emulate. Everyone has a different superpower, and you build your toolkit by learning from each person. A strong group of player-coaches raises the entire team&#8217;s game because people learn from one another.</p><h3>As the product life cycle accelerates, how do you maintain focus on selecting the right customer problems to solve?</h3><p>As building becomes cheaper, teams can fall into the trap of testing every idea or, worse, shipping it just to see what happens. Faster production does not eliminate the costs of distracting customers, fragmenting the product experience, or consuming organizational resources.</p><p>I use two mental models to counter that. First, separate the problem from the solution. Choosing which customer problem to solve should be evidence-based and relatively durable. The solution can be much more fluid; teams should test, learn, and pivot without abandoning the problem. Stress-testing the problem itself helps keep teams anchored.</p><p>Second, distinguish speed from velocity. Speed is how quickly you move; velocity is how quickly you move in a specific direction. A good product strategy establishes that direction. Strategy cycles are shortening as AI changes the market rapidly, but strategy itself is more important, not less.</p><p>AI should increase the number of solutions we explore, not the number of disconnected problems we pursue. Product leaders must maintain conviction about the destination while giving teams and tools the freedom to find a faster path.</p><h3>What does LogRocket do?</h3><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/?substack">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[A Real-Time Look Inside a CPO’s First 30-60-90 Days | Maya Keekattil, CPO (Cin7)]]></title><description><![CDATA[Cin7's new Chief Product Officer on why your first 30 days should be for listening, how to find quick wins without pulling out load-bearing beams, and why every mandate should go back to the customer.]]></description><link>https://stories.logrocket.com/p/real-time-look-inside-cpo-first-30-60-90-days-maya-keekattil</link><guid isPermaLink="false">https://stories.logrocket.com/p/real-time-look-inside-cpo-first-30-60-90-days-maya-keekattil</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 13 Oct 2026 13:58:47 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/f705e106-be96-4f6b-bd0e-123a76485ef4_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div id="youtube2-yc7NjtCdK8M" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;yc7NjtCdK8M&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/yc7NjtCdK8M?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M">YouTube</a> | <a href="https://open.spotify.com/episode/0bt8KdnqDGjYRAMhQY1EAs">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/a-real-time-look-inside-a-cpos-first-30-60-90/id1733103005?i=1000796130393">Apple</a></strong></em></p></div><p>Most advice about a leader&#8217;s first 90 days is written in hindsight, after the dust has settled and the mistakes have been solved. This episode is different &#8212; when we recorded this, Maya Keekattil was ending her third week as CPO at Cin7. She was working through her first 30 days in real time.</p><p>Maya started as a software engineer and DBA before she moved into project management, taking on product leadership at Smarsh, where she spent nearly five years as VP of Product Management. Now she&#8217;s learning a new vertical at a PE-backed company, where the board, leadership, and her team all have high expectations.</p><p>In this episode, Maya shares:</p><ul><li><p>Why every product leader&#8217;s first 30 days should be reserved for listening and learning, and the three questions she asks in every one of her cross-functional one-on-one meetings</p></li><li><p>How to find fast operational wins without accidentally ripping out something load-bearing</p></li><li><p>How she&#8217;s restructuring product teams around outcomes, not products, in the AI era</p></li></ul><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. The first 30 days are for listening and learning</h2><p>Maya understands that new leaders feel pressure to bring change quickly and to prove results and ROI &#8212; this is especially true in the PE world. But Maya emphasizes the importance of learning how things work before you shake anything up. In three weeks, she scheduled 51 one-on-ones across her org, and she&#8217;s aiming for 75. Those meetings go beyond her own team, into sales, marketing, finance, engineering, and operations. She told us she started booking conversations a few weeks <em>before</em> her official start date.</p><p>Every conversation comes back to three questions: What&#8217;s working? What&#8217;s not working? And what do you expect product to do for you?</p><blockquote><p>&#8220;The company has been here for a decade and more. So there has been stuff that is successful. Coming in and thinking that I&#8217;m going to change it all is probably a big mistake a lot of leaders make.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Product means something different at every company, and you can&#8217;t meet expectations you don&#8217;t understand. Before you change anything, learn what each cross-functional partner expects from product. That should be your first scorecard.</p><div><hr></div><h2>2. Patch the holes in the boat first </h2><p>How do you get quick results without pulling out a beam that&#8217;s holding up the building? Maya&#8217;s answer is to start with operational efficiency: the small process gaps that slow everyone down.</p><p>Often, she says, other people noticed these gaps long ago. They didn&#8217;t fix them because they lacked the operational expertise or they simply didn&#8217;t have the authority to do so. A new leader with fresh eyes and the mandate to act can rally people quickly around the right fixes.</p><blockquote><p>&#8220;Most of the time, others would have observed it. They probably either lacked the expertise to do something about it, were hesitant to do it, or their role did not allow them to do it. Where I stand today, my role allows it, I have the authority to do it, and I come with this fresh perspective.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Your fastest credible wins usually aren&#8217;t new features. They&#8217;re process fixes that your peers have quietly wanted for a long time. Ask where things get stuck, fix what you can in the first 30 to 60 days, and you&#8217;ll build the trust you need for bigger bets down the line.</p><div><hr></div><h2>3. Strategy can&#8217;t wait for onboarding to end: Don&#8217;t wait until day 61 to think strategically </h2><p>Many 30-60-90 plans treat strategy as a phase you reach once onboarding is finished. Maya warns that this creates a gap between your arrival and your first real impact. Strategic changes need alignment, investment, and often board involvement, so they take longer to show results. </p><blockquote><p>&#8220;I don&#8217;t think it&#8217;s a switch. You start rolling on this innovation and strategy perspective even through the 30/60, but it is not apparent to the cross-functional teams.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Run operational and strategic work at the same time. Fix things in the open, and start the slower alignment conversations with your executive peers in the background.</p><div><hr></div><h2>4. Every mandate should go back to the customer </h2><p>Every new leader arrives with a mandate. Maya argues that the mandate itself is too short-sighted to plan around. Her starting point is two things: why customers buy, and why they leave.</p><blockquote><p>&#8220;We are not looking at just numbers. We are looking at why is the customer buying us today and why is the customer leaving us today. Just those two things will get you to the gist of where you are succeeding versus where you are not.&#8221;</p></blockquote><p>She says even customers who have already left are worth talking to:</p><blockquote><p>&#8220;Even a customer who has left is very happy to talk to you, and they spend the time because they are invested. They&#8217;ve used your product.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> When the board hands you a target, don&#8217;t work backward from the number. Work forward from churn. In Maya&#8217;s experience, building around the customer&#8217;s view beats the mandate. Building <em>to</em> the mandate rarely does.</p><div><hr></div><h2>5. Structure the team around outcomes, not products </h2><p>Maya saw Y2K and the move to the cloud, and she says <em>now</em> is the fastest she has ever seen technology move &#8212; thanks to AI pushing companies to organize around outcomes, which shows up in everything from team structure to outcome-based pricing models.</p><p>Instead of splitting PMs by product line, Maya thinks in layers. Platform-focused technical PMs build the foundation. AI product managers operationalize AI across the organization. Growth PMs are tied to customer outcomes.</p><p>Before assigning anyone to a role, she starts with an honest look at each person&#8217;s strengths. And on whether to hire for new skills or develop the team she has, Maya puts a high value on domain expertise:</p><blockquote><p>&#8220;Even if you hire for a talent gap, you need to bring someone up to speed on the domain. So I would rather trust the teams that you have on the domain expertise and build that up. You are <em>influencing</em> and not just using your authority. That goes a long way.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Before you reorganize for AI, separate true talent gaps from skills your current team can learn. Hire for the first, then invest in the second. Domain knowledge is the hardest thing to buy and the slowest to rebuild.</p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;pp=0gcJCWMAwfN6Pr3D">00:00</a> Introduction<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;t=127s&amp;pp=0gcJCWMAwfN6Pr3D">02:07</a> Maya&#8217;s path from software engineer to chief product officer<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;t=218s">03:38</a> Why your first 30 days should be for listening and learning<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;t=370s">06:10</a> Three questions to ask in every cross-functional one-on-one<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;t=479s&amp;pp=0gcJCWMAwfN6Pr3D">07:59</a> Finding quick operational wins without breaking what works<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;t=669s">11:09</a> When to shift from operational fixes to long-term strategy<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;t=969s">16:09</a> Defining success by why customers buy and why they leave<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;t=1337s">22:17</a> How AI is reshaping product team structure around outcomes<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;t=1637s">27:17</a> Developing the team you inherited vs. hiring for talent gaps<br><a href="https://www.youtube.com/watch?v=yc7NjtCdK8M&amp;t=1769s&amp;pp=0gcJCWMAwfN6Pr3D">29:29</a> Conclusion</p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/mayakeekattil/">Maya&#8217;s LinkedIn</a></p></li><li><p><a href="https://www.cin7.com/">Cin7</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at LogRocket.com.</p>]]></content:encoded></item><item><title><![CDATA[PM Was Scrumful, Not Agile: Why AI Can Give Product Its Real Job Back | Naomi Lariviere, CPO (ADP)]]></title><description><![CDATA[ADP's Chief Product Officer on why PMs got "Scrumful, not Agile," what happened when her dev team tried to ship without product, and why storytelling and judgment matter more than ever.]]></description><link>https://stories.logrocket.com/p/pm-scrumful-not-agile-why-ai-can-give-product-real-job-back-naomi-lariviere</link><guid isPermaLink="false">https://stories.logrocket.com/p/pm-scrumful-not-agile-why-ai-can-give-product-real-job-back-naomi-lariviere</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 06 Oct 2026 13:23:06 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/68e69949-8452-4bc3-a403-47e06bb0430c_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div id="youtube2-__yy43aDuHw" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;__yy43aDuHw&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/__yy43aDuHw?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw">YouTube</a> | <a href="https://open.spotify.com/episode/5gXhD2Q4d27A7wZmL11tRn">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/your-ai-doesnt-have-a-feature-problem-it-has/id1733103005?i=1000792193319https://podcasts.apple.com/us/podcast/pm-was-scrumful-not-agile-why-ai-can-give-product-its/id1733103005?i=1000793445582">Apple</a></strong></em></p></div><p>AI has made building faster and cheaper than ever. That leaves product leaders with an uncomfortable question: if engineering isn&#8217;t the bottleneck anymore, what is? And do we still need product managers?</p><p>Today&#8217;s guest has a clear answer. Naomi Lariviere is Chief Product Owner at ADP, where she oversees a roughly $4 billion mid-market portfolio for a company that pays one in six Americans. Her career path is unusual for a CPO: she started as a business systems analyst in financial services, then stepped away from product to run an IT PMO, professional services, and even pricing and sales ops before coming back.</p><p>Her take: much of the PM job over the past decade drifted toward tickets and standups instead of customers and markets. AI is finally a chance to hand that work off and get back to the real job.</p><p>In this episode, Naomi shares:</p><ul><li><p>What happened when ADP&#8217;s dev team tried to deliver without a product team, and why they came back</p></li><li><p>How ADP studied what its PMs actually do all day to find the busywork AI can take off their plates</p></li><li><p>Why she&#8217;s skeptical of the push to turn every PM into a product builder who ships code</p></li><li><p>How ADP keeps innovating when a single mistake could mean thousands of people don&#8217;t get paid</p><div><hr></div></li></ul><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. The experiment: Shipping without product </h2><p>At ADP, the question of whether product was still necessary didn&#8217;t stay theoretical. When the idea came up that product could be eliminated, Naomi handed the dev team the challenge and told them to figure out how to deliver something on their own.</p><p>It didn&#8217;t take long for the gaps to show.</p><blockquote><p>&#8220;Ultimately we ended up, &#8216;Oh, wait. We need the product person to tell us, like, well, what are the rules? What is the compliance regulation?&#8217;&#8221;</p></blockquote><p>Naomi&#8217;s explains why this happened:</p><blockquote><p>&#8220;AI makes building cheaper, but it doesn&#8217;t make choosing things any easier.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Speed without direction isn&#8217;t progress. As AI compresses build time, the scarce resource becomes knowing which problem to solve, under which constraints, for which customer. If your org is debating whether product is still needed, the fastest way to answer is to look at what breaks when nobody owns the rules, the regulations, and the &#8220;why.&#8221;</p><div><hr></div><h2>2. How PMs got &#8220;Scrumful, not Agile&#8221;</h2><p>Early in her career, Naomi spent her time with clients and prospects, doing market research and studying the competitive landscape. Then Agile arrived, and the expectations quietly shifted. PMs were now expected to write every requirement, run the backlog, and direct each sprint.</p><p>The cost of that shift was everything PMs stopped doing: </p><ul><li><p>Talking to clients</p></li><li><p>Understanding the market</p></li><li><p>Reading the macroeconomic picture</p></li></ul><p>And that&#8217;s exactly why AI looks like a threat to some. If the job is writing PRDs and user stories, AI can do it. </p><p>But that was never the actual job:</p><blockquote><p>&#8220;What product actually is doing is synthesizing multiple inputs and then going, &#8216;Here&#8217;s the best opportunities that we have to solve a problem, and here&#8217;s the ones that are gonna make you money.&#8217;&#8221;</p></blockquote><p><strong>Product takeaway:</strong> If your team&#8217;s value is measured by how well it runs ceremonies, it&#8217;s vulnerable. Look honestly at where your PMs&#8217; hours go. Process should exist to drive decisions, and every recurring meeting should earn its place by unlocking one. Three standups across three teams is busywork, not product management.</p><div><hr></div><h2>3. Innovating when a mistake means someone doesn&#8217;t get paid</h2><p>Startups can push code multiple times a day. ADP, with over 1.1 million clients, operates with much higher risk. They pay 1 in 6 Americans, and as Naomi says:</p><blockquote><p>&#8220;When you think about somebody&#8217;s pay, they could be counting on that check to make their rent or, you know, pay for a vacation, or they&#8217;re getting married and they need to put a deposit on an event space. You know, if we do something that means they miss that, that is, that&#8217;s not okay.&#8221;</p></blockquote><p>That doesn&#8217;t mean ADP doesn&#8217;t move. It means it moves deliberately:</p><blockquote><p>&#8220;We innovate actually all the time, but we do it very thoughtfully and in a very controlled fashion.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> &#8220;Move fast&#8221; isn&#8217;t a universal virtue. The right release cadence depends on the cost of being wrong for your users. In high-stakes domains like payroll, finance, and healthcare, the product team&#8217;s job includes keeping the user&#8217;s real-world consequences at the center of every innovation decision, and building the controls that let you ship confidently.</p><div><hr></div><h2>4. Product managers, not product builders</h2><p>There&#8217;s a growing push to turn every PM into a builder who ships code with AI. Naomi thinks that misreads what the role is for.</p><blockquote><p>&#8220;Yes, I do know how to code, but do you trust me at launching feature and functionality into a platform like one of ADP&#8217;s? Probably you shouldn&#8217;t do that.&#8221;</p></blockquote><p>The skills that actually make a PM effective are the ones that coding doesn&#8217;t touch. Interviewing a client without introducing your own bias. Understanding what pushes someone to buy. And above all, telling the story:</p><blockquote><p>&#8220;When someone asks me what would be the one surprising thing about my job that most people don&#8217;t understand, it&#8217;s the storytelling, it&#8217;s the influencing that we do. And, you know, coding doesn&#8217;t help me do that.&#8221;</p></blockquote><p>Naomi does expect roles to become more fluid, but she wants developers doing what they&#8217;re great at and product people doing what they&#8217;re great at.</p><p><strong>Product takeaway:</strong> Before requiring that every PM become a builder, ask what you&#8217;d lose. The PMs everyone in an org seeks out, from sales to engineering to execs, are rarely the best coders. They&#8217;re the ones who read the room, carry empathy for the user, and articulate why something matters. Protect and invest in those skills; they&#8217;re becoming the core of the job.</p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;pp=0gcJCWMAwfN6Pr3D">00:00</a> Introduction<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=144s">02:24</a> Naomi's path from business analyst to ADP product leader<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=259s">04:19</a> Is AI the end of product management?<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=372s&amp;pp=0gcJCWMAwfN6Pr3D">06:12</a> AI makes building cheaper, not choosing easier<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=502s">08:22</a> How product teams became "Scrumful, not Agile"<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=778s">12:58</a> Why ADP has to innovate carefully at its scale<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=932s">15:32</a> Process with a purpose vs. process for its own sake<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=1050s">17:30</a> Should product managers become product builders?<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=1222s">20:22</a> ADP's jobs-to-be-done study of the PM role<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=1333s">22:13</a> Fluid roles and why less technical PMs can deliver more<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=1543s">25:43</a> Storytelling, empathy, and the human skills AI can't replace<br><a href="https://www.youtube.com/watch?v=__yy43aDuHw&amp;t=1703s">28:23</a> Conclusion</p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/naomilariviere/">Naomi&#8217;s LinkedIn</a></p></li><li><p><a href="https://www.adp.com/">ADP</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at LogRocket.com.</p>]]></content:encoded></item><item><title><![CDATA[Your AI Doesn't Have a Feature Problem. It Has an Adoption Problem. | Charanya Kannan (Navan)]]></title><description><![CDATA[Navan's Charanya Kannan on building AI distribution-first, building more than just features that won&#8217;t get adopted, and how Navan grew its margins while the rest of SaaS braces for compression.]]></description><link>https://stories.logrocket.com/p/your-ai-doesnt-have-feature-problem-has-adoption-problem-charanya-kannan</link><guid isPermaLink="false">https://stories.logrocket.com/p/your-ai-doesnt-have-feature-problem-has-adoption-problem-charanya-kannan</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 29 Sep 2026 13:13:17 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/2404928b-27ff-473c-9b31-5d8824482527_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p></p><div id="youtube2-wMJeMW_paC0" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;wMJeMW_paC0&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/wMJeMW_paC0?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=wMJeMW_paC0">YouTube</a> | <a href="https://open.spotify.com/episode/2i8YgKzbJ2ZEySDnCFRNo3">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/your-ai-doesnt-have-a-feature-problem-it-has/id1733103005?i=1000792193319">Apple</a></strong></em></p></div><p>Every product team is shipping AI. But for many companies, very few users are adopting it. The people who try the feature love it, and adoption stalls anyway.</p><p>Charanya Kannan thinks that&#8217;s because teams obsess over <em>building</em> AI and barely think about <em>adoption</em>. As VP and GM of Navan Anywhere, she puts Navan&#8217;s AI travel and expense tools inside Slack, Teams, and Gemini, where people already work, instead of asking them to open yet another app. She joined Navan when it was still called TripActions and doing under $50M in revenue. Now it&#8217;s well on its way to $1B!</p><p>Her core belief: if customers have to work harder to use your AI, it isn&#8217;t actually better.</p><p>In this episode, Charanya shares:</p><ul><li><p>How Navan Anywhere was built distribution-first, bringing booking and expenses into the tools people already use every day</p></li><li><p>Why Navan&#8217;s margins went up in the AI era, while the rest of SaaS braces for compression</p></li><li><p>And why she believes PMs who mostly manage process and Jira tickets will fade away, while those who can actually drive user benefit and business outcomes will matter more than ever</p></li></ul><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. Building is easy. Getting users to adopt is the real problem.</h2><p>These days, you can create an  AI prototype in about a weekend. But getting anyone to actually use it? That&#8217;s where most teams get stuck. </p><p>Charanya&#8217;s fix is to skip the &#8220;please come try our shiny new thing&#8221; step entirely. Navan Anywhere lives inside Slack, Teams, and Gemini, where people already spend their day. They don&#8217;t have to download a new app or form a new habit because, as she says: </p><blockquote><p>&#8220;If they have to put in extra effort to do something because AI is superior, then AI is not really superior.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Plan for adoption <em>as</em> you plan the feature. The easiest product to adopt is the one that shows up where your users already are.</p><div><hr></div><h2>2. Not everything needs to be a chatbot</h2><p>Six months ago, Charanya told her designers to start brushing up on evals, because chat was going to eat everything. Then users actually tried picking a hotel from five paragraphs of text. It turned out that people booking hotels wanted to see photos of the hotel.</p><p>So, Navan mixes and matches. Chat handles the &#8220;here&#8217;s what I want&#8221; part and the UI handles the &#8220;let&#8217;s compare&#8221; part, while payments stay firmly on the reliable deterministic rails.</p><blockquote><p>&#8220;The real judgment or discernment is: Where do you bring in AI? Where does it make sense to bring in your agentic systems versus where do you fall back to your deterministic systems?&#8221;</p></blockquote><p><strong>Product takeaway:</strong> &#8220;AI everywhere&#8221; isn&#8217;t a strategy. Use AI where it removes friction, use UI where people need to see things, and be skeptical about AI related to anything that moves money.</p><div><hr></div><h2>3. The two kinds of PMs &amp; why nobody&#8217;s irreplaceable because they write great Jira tickets</h2><p>Charanya has mentored dozens of PMs, and she sees two types: Those who spend their days running standups and shuttling messages between sales and engineering, and those who ask why the team is building something at all, and whether it&#8217;ll still matter in five years.</p><p>Only one of those jobs survives the AI era. </p><p>Navan runs a travel business doing over $3B in bookings with fewer than five PMs, and engineers write a lot of their own tickets.</p><blockquote><p>&#8220;Everybody knows there&#8217;s a problem, but how do you frame that problem? Give the problem eyes and ears and a figure that people can go after and solve it&#8230; If you&#8217;re just thinking about processes 80% of your time, that&#8217;s not what we want.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Process should be about 20% of the job. If it&#8217;s 80%, AI is coming for that part first.</p><div><hr></div><h2>4. How Navan&#8217;s margins went up in the AI era</h2><p>While most of SaaS is sweating over inference bills, Navan&#8217;s margins are growing. The trick is letting AI do what it&#8217;s good at: it now resolves about half of support conversations, up from a third back in the pre-LLM days.</p><p>The truly messy stuff still goes to human travel counselors. </p><blockquote><p>&#8220;When you&#8217;re stuck in a plane for five hours on the tarmac&#8230; no AI can come and solve the problem for you.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Let AI handle the repetitive stuff, save your humans for the hard stuff, and treat your model choices like the budget decisions they are.</p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=wMJeMW_paC0">00:00</a> Introduction<br><a href="https://www.youtube.com/watch?v=wMJeMW_paC0&amp;t=257s">04:17</a> Why AI features don't always get adopted<br><a href="https://www.youtube.com/watch?v=wMJeMW_paC0&amp;t=390s">06:30</a> Building Navan Anywhere distribution-first<br><a href="https://www.youtube.com/watch?v=wMJeMW_paC0&amp;t=613s">10:13</a> When to use AI vs. deterministic systems<br><a href="https://www.youtube.com/watch?v=wMJeMW_paC0&amp;t=721s">12:01</a> Conversational cognitive load and why good UI still matters<br><a href="https://www.youtube.com/watch?v=wMJeMW_paC0&amp;t=957s">15:57</a> Evals, distilled models, and the AI reliability pyramid<br><a href="https://www.youtube.com/watch?v=wMJeMW_paC0&amp;t=1214s&amp;pp=0gcJCWMAwfN6Pr3D">20:14</a> Product owners vs. product managers in the AI era<br><a href="https://www.youtube.com/watch?v=wMJeMW_paC0&amp;t=1570s">26:10</a> How Navan is growing margins while SaaS faces AI compression<br><a href="https://www.youtube.com/watch?v=wMJeMW_paC0&amp;t=1900s">31:40</a> Conclusion</p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/meetcharanya/">Charanya&#8217;s LinkedIn</a></p></li><li><p><a href="https://navan.com/">Navan</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[Leader Spotlight: Why trust is the real UX problem, with Angélique Wynants]]></title><description><![CDATA[Ang&#233;lique Wynants is Senior Director, End User at EliseAI, where she&#8217;s building the company&#8217;s first dedicated end-user experience for renters on a platform designed primarily for property management companies.]]></description><link>https://stories.logrocket.com/p/leader-spotlight-angelique-wynants</link><guid isPermaLink="false">https://stories.logrocket.com/p/leader-spotlight-angelique-wynants</guid><dc:creator><![CDATA[Katie Schickel]]></dc:creator><pubDate>Tue, 29 Sep 2026 07:03:03 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!lDB4!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><em><span>Ang&#233;lique Wynants is Senior Director, End User at EliseAI, where she&#8217;s building the company&#8217;s first dedicated end-user experience for renters on a platform designed primarily for property management companies. Her career has centered on designing for people navigating high-stakes, unfamiliar systems. At Nubank, she built for people getting a bank account for the first time in Mexico City, and at Comun, she led operations for a digital bank built specifically for undocumented immigrants. She studied economics at the London School of Economics.</span></em></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!lDB4!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!lDB4!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!lDB4!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!lDB4!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!lDB4!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!lDB4!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png" width="895" height="597" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/dd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:597,&quot;width&quot;:895,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1347892,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://stories.logrocket.com/i/217115718?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!lDB4!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!lDB4!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!lDB4!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!lDB4!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd49979f-06da-43cb-b8ef-5dc85aff0e5c_895x597.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em><span>In this conversation, Ang&#233;lique shares the three principles she returns to when designing for the underserved and never-before-served: education, simplicity, and support. She talks through the assumptions that trip up well-intentioned product teams, the community events that replaced traditional user research, and a security feature that tanked adoption the moment her team stopped thinking from the user&#8217;s point of view. She also dives into what it&#8217;s like building EliseAI&#8217;s first B2C product inside a B2B company and where AI belongs in a support experience built for people with the least room for error.</span></em></p><div><hr></div><h2><span>Building for the underserved</span></h2><h3><span>Who are the end users you&#8217;ve built for at Nubank, Comun, and now EliseAI?</span></h3><p><span>In the first part of my tech career, at Nubank and Comun, I was building for end users who were new to financial services and digital financial services. These were people who were getting a credit card or a bank account for the first time, or sending a remittance to a different country for the first time, or who were unfamiliar with using apps for banking needs. At Elise, I look after the experience of those going through the experience of finding a home, which is, I&#8217;d say, equally critical as financial services.</span></p><p><span>Here, it&#8217;s making this experience intuitive for the full spectrum of end users. This could be affordable housing applicants who are going through the Elise flow for the first time trying to apply for a home, or who have never used technology for any of their housing needs. Their needs are similar &#8212; both are going through a highly stressful moment in their life where they&#8217;re sending money to someone, getting a loan, applying for an apartment. It&#8217;s been a nice thread in my career, building for that subset.</span></p><p><span>Comun was specifically focused on undocumented immigrants. It was a digital bank for undocumented immigrants. It served everyone, really, but because the product enables people to get a bank account without a Social Security Number, it particularly catered to undocumented immigrants. The app was also fully in English and Spanish, so it catered even more to Latino undocumented immigrants. At Nubank, back in Mexico City, they were predominantly people from Mexico but who had never had a bank account before and previously were maybe saving under their mattress, and for the first time wanted to use more formal financial services.</span></p><h2><span>Three guiding principles: Education, simplicity, and support</span></h2><h3><span>What principles matter most when you&#8217;re designing for users with limited experience or comfort with technology?</span></h3><p><span>Typically, I always keep three principles in mind when I build for this type of end user.</span></p><p><span>The first one is education. I don&#8217;t assume any prior knowledge of the product. This could be explaining how interest rates work on a credit card, which you and I would know, but that&#8217;s not at all intuitive to someone who&#8217;s never owned a credit card before. Or being transparent about how income requirements work in housing, like someone submitting their income and not understanding why they&#8217;re below certain thresholds. Things like linking FAQs, educational videos, a UI wording that is educative is very, very important.</span></p><p><span>The second one is simplicity. Less is more. The UI needs to be really simple, straight to the point. I always try to make sure we minimize the amount of screens to complete a flow, that the browsing is simple, that every step is over-explained in simple terms &#8212; again, not assuming that this person has ever gone through a flow like this before.</span></p><p><span>The third one, which I think is really important and often overlooked, is support. I&#8217;ve done a lot of support work also in my career, so it&#8217;s very important to always give them a way to reach out for help. The first layer of support is increasingly AI, obviously, but they should never be stuck in the loop of either the technology not being able to help them, or not knowing where to look for help. This is very important, that there is an option for support, AI or human.</span></p><h2><span>Where product teams go wrong</span></h2><h3><span>What do product teams tend to get wrong when they design for users whose financial circumstances, language skills, or digital literacy differ from their own?</span></h3><p><span>This is one of the most surprising aspects of building for low-income or low-tech-sophistication users. Across teams that I&#8217;ve led and worked with, I always see two assumptions that people repeatedly make. The first: we tend to assume a lot of previous knowledge about a product or process, and someone who&#8217;s never used technology before, or is new to a system, a country or a process, may not necessarily know. One example is when applying for an apartment, what is a guarantor? When I moved to the U.S., I wasn&#8217;t familiar with what that meant, why I needed it, what they needed to show &#8212; whereas that&#8217;s a very intuitive concept for anyone who&#8217;s applied for an apartment previously. Never assume previous knowledge.</span></p><p><span>The second: teams assume a level of trust in technology that isn&#8217;t actually adapted to the realities of the end user. There&#8217;s a political scientist I really like called Francis Fukuyama, and he writes about low-trust and high-trust economies &#8212; how certain countries or communities operate in a fundamentally low-trust environment. Their trust circle is smaller; they tend to distrust larger organizations, institutions, new technologies. But people working in tech inherently trust tech. They don&#8217;t necessarily think that if you&#8217;re asking someone to input their bank information on a mobile, this person is going to think they&#8217;re being scammed, because that&#8217;s not going to be their first intuition. But that is the first intuition of a lot of people living in lower-trust communities, in which they often get scammed, for instance. The credit card is an obvious one &#8212; there&#8217;s a lot of social proof that we need to build to increase product adoption. In FinTech there&#8217;s a lot of things we can do. I think in housing as well, there&#8217;s a lot of things we can do to build trust, but that&#8217;s very often overlooked.</span></p><h3><span>Can you walk us through a product decision that changed once your team better understood how the user was actually navigating their experience?</span></h3><p><span>One example from one of my previous jobs is that we wanted to make the bank account connection more secure. We wanted to enable people to connect their primary bank account so that they could pull the money from the other bank account and then make payments. One way of doing that, which is quite common in the U.S. banking sector, is what we call micro verification. It says, &#8220;Hey, I sent you two cents on your primary bank account.&#8221; Now you go to the primary bank account, you see the two cents, and then you have to enter the amount. Most people working in tech probably have gone through that experience before, using digital banking services. We implemented that, and our product adoption completely dropped. People were very confused. They didn&#8217;t know how to use it.</span></p><p><span>They distrusted the fact that they had to enter their other bank account information. They weren&#8217;t sure how to go back and forth from one app to the other. So we had to quickly find another solution: building a very sophisticated risk score, so we&#8217;d only add this friction for very high-risk users who may be fraudulent, and find other ways to do verification for everyone else. Most of the undocumented immigrants we were working with &#8212; this was their first time using a digital service, and now we were asking them to open three apps at once and do something called micro verification. It just wasn&#8217;t viable. For the ones who did have to continue with that process, we also built educational videos &#8212; step-by-step, why it&#8217;s important, and how long it will take.</span></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive new posts every week.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h2><span>Getting inside the user&#8217;s world</span></h2><h3><span>Is there anything you do in particular to help your teams get in the mindset of the user?</span></h3><p><span>I think user research is really important. There is traditional user research, but there is also non-traditional user research &#8212; going to the places where the target user lives and works or applies for an apartment, talking to them, seeing them using apps and technology, understanding what they&#8217;re afraid of. It also, frankly, requires a level of empathy that is really important. When building teams that build for the end user, it requires a different and, I think, higher level of empathy to be able to put yourself in someone else&#8217;s shoes, in a life that you have never lived and probably will never live.</span></p><h3><span>What have you done to get meaningful feedback when traditional user research doesn&#8217;t work?</span></h3><p><span>This is interesting because when I started at Comun, which is the digital bank that builds for undocumented immigrants, I tried conducting regular user research calls. I would reach out to our users and try to schedule a call with them. Friday, 2:00 PM &#8212; they would say, &#8220;That&#8217;s great.&#8221; I would send a calendar invite, and then I would connect at Friday, 2:00 PM, and no one would connect. This would happen repeatedly. At first, I didn&#8217;t understand why. So then what I tried to do is just call their number, a few different times throughout the day &#8212; obviously, only when I had their approval to reach out to them.</span></p><p><span>I realized that they ended up picking up at different times of the day, and were very happy to talk to me, but they just happened to be in their car commuting, or they&#8217;d just gotten home. A lot of them don&#8217;t have traditional working hours &#8212; different shifts, night shifts, construction sites, cleaning services &#8212; and they don&#8217;t know where they&#8217;re going to work the next day, or what time. So even though they wanted to help and wanted to talk to us, we couldn&#8217;t stick to a very specific schedule. What I realized is that what we really had to do is go and meet them where they are, in the spaces where they live, work, or hang out.</span></p><p><span>At Comun, because a lot of our users were from the Latino population, we would go to local community events and have stands, and talk to people there &#8212; &#8220;Hey, have you heard of Comun? Tell us more about how you send money back to Colombia. Tell us more, what bank do you use? Do you have a bank, or do you use your husband&#8217;s bank?&#8221; Because a lot of the time we also realized they share the accounts, for instance. How do they get paid? Those informal conversations, in person and at community events, ended up being the most relevant for knowing what to build and what the pain points were &#8212; a lot more than phone conversations, which were very difficult to get.</span></p><h2><span>Designing for high-stakes decisions: Money and housing</span></h2><h3><span>EliseAI is a B2B platform for property management companies and you&#8217;ve been brought in to build the B2C end user experience for the renter. How do you balance the needs of both, especially when they have different priorities?</span></h3><p><span>This is a very interesting question because there are very, very tough trade-offs between prioritizing B2C and B2B &#8212; something I&#8217;ve experienced for the first time joining Elise, because it&#8217;s the first time I&#8217;m joining a company that&#8217;s B2B. When working at a company that&#8217;s purely B2C, the end user is the client; the revenue of the company depends on whether the end user likes the product or not. That&#8217;s very different from B2B, in which we&#8217;re selling to businesses, and the client is not the end user. Elise&#8217;s platform is built for B2B, so they remain the main client. So when there are very urgent B2C issues, this is always prioritized, because in a way it can also impact our client. If someone misses out on a unit &#8212; Elise works with some of the largest property managers in the country, some with buildings in very remote areas &#8212; that unit can be vacant for a couple of months. In that way, their issues overlap.</span></p><p><span>But a lot of the property managers also have feature requests that have no impact whatsoever on the end user. That&#8217;s when it sometimes conflicts: should we improve the end user-facing platform, making the flow more educational or simple, or improve a feature within the B2B-facing app so property managers can more easily understand what&#8217;s pending on their side? That&#8217;s difficult, because the property managers are the clients &#8212; that weighs heavily in the prioritization.</span></p><h2><span>AI, education, and trust</span></h2><h3><span>When users need to understand things like credit scores, income requirements, or why an application was denied, where should education live within the product experience? And how does AI adoption differ across your end users?</span></h3><p><span>This has changed quite a bit with AI and is still changing. Previous to AI, when I was at Nubank, there was a typical help section, which would have the FAQs, educational videos, and we would use LLMs to maybe customize what FAQ would show up first. We would also link FAQs and videos throughout the UI, which is still something that I try to push teams to do now as well.</span></p><p><span>But now with AI, the new and better way to do it, in my opinion, is to have an embedded AI assistant that is constantly visible throughout, desktop or mobile. Instead of going to search the help section and having to press a back button and then coming back to the step where they&#8217;re struggling, there&#8217;s always an AI bubble that&#8217;s there, like, &#8220;Hey, I&#8217;m not sure what that step is, or what ID can I upload? or why is this failing?&#8221; And so they have a personal assistant that&#8217;s helping them throughout the process. Another interesting thing to explore too is other channels outside of the product &#8212; WhatsApp works very well for some communities. If they submit a support ticket in the app, or they&#8217;re talking to the AI and they exit the app, they just won&#8217;t go back to it. On WhatsApp, they actually respond immediately.</span></p><p><span>It&#8217;s a mix. The trust issue is definitely there, but what&#8217;s interesting is that even when we have humans responding, we are now being asked if it&#8217;s AI &#8212; the humans also have a certain tone of voice. It&#8217;s hard now to differentiate when it&#8217;s a human or when it&#8217;s AI. I would say for lower income populations, or even immigrants who don&#8217;t necessarily speak English well, long texts don&#8217;t work at all. So it&#8217;s either a short text, or some of them want to call. What I&#8217;ve done in the past is making sure we send videos &#8212; a series of how-to videos on YouTube &#8212; so when someone&#8217;s like, &#8220;I still don&#8217;t understand, that&#8217;s too long,&#8221; we just send a link, and then they get it that way.</span></p><p><span>My philosophy when it comes to support is that they should never be stuck in a loop of no help. So we have a support team &#8212; now at Elise it&#8217;s 7:00 AM to midnight &#8212; where if the first layer of AI is not able to help solve the issue, it escalates to a team that is trained to respond to those questions.</span></p><h3><span>How do you apply your guiding principles&#8212;education, simplicity, and support&#8212;when deciding between using an AI assistant versus keeping a human in the loop?</span></h3><p><span>Education is educating the AI to use an educational tone. I actually feed the AI a tone of voice, which is the same one that I use to train the support team: empathetic, straight to the point, avoiding extra words so that the text is not too long. I literally have dos and don&#8217;ts, such as do not say, &#8220;Hi, thank you so much for contacting us.&#8221; No, it&#8217;s, &#8220;Hi, I saw that this is your issue. This is how we&#8217;re going to solve it. This is what you can do next,&#8221; in a very empathetic tone, but we don&#8217;t want three paragraphs. So it&#8217;s training the AI to be educational, and even the prompt assumes no prior knowledge, and explains in a simple, straight-to-the-point way. That also works for simplicity and support. The AI needs to be escalating when it doesn&#8217;t understand or it&#8217;s not able to help the user, instead of hallucinating or giving answers that aren&#8217;t there.</span></p><p><span>Even though it&#8217;s the same underlying technology, the prompts we use for AI responses to B2C users are completely different from the prompts we use for our B2B customers or our internal tech team. The knowledge is different, the guidance is different, the prompts are different.</span></p><h3><span>As AI moves into essential services, how do you make sure it removes barriers rather than creating new ones?</span></h3><p><span>I think the knowledge feeding the AI is really crucial to make sure AI is an enabler rather than a barrier. This technology is very, very powerful and we can model it however we&#8217;d like. The type of knowledge that educates the user is truly important, so that it becomes an educational tool, rather than something that&#8217;s confusing or making it even more complex for them to understand.</span></p><p><span>I also think about enabling AI to decision, as opposed to just a chatbot. AI can make faster decisions. So once it takes action, it contributes to removing barriers, which are actually often human bottlenecks &#8212; like decisioning on an application or on a credit card, or making personal information changes to a bank account that a bank clerk would otherwise make, or a few hours waiting at a bank branch. The speed of AI itself can remove barriers rather than creating new ones. And, obviously, we need guardrails in place. Having a very strong product QA in place is really important &#8212; I always have a team doing product QA and human QA. And we always have a way for the end user to escalate any issues, so that if the AI isn&#8217;t being an enabler and rather is being a barrier, we can catch that and iterate on it.</span></p><h3>What does LogRocket do?</h3><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/?substack">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[Building AI Products for the Other 95% | Brian McMullin, SVP Product (Network Solutions)]]></title><description><![CDATA[Brian McMullin visits to talk why small business owners don't want another prompt box, what to do about inference costs before token pricing settles, and why churn is usually about bad basics.]]></description><link>https://stories.logrocket.com/p/building-ai-products-other-95-brian-mcmullin</link><guid isPermaLink="false">https://stories.logrocket.com/p/building-ai-products-other-95-brian-mcmullin</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 22 Sep 2026 13:19:08 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/4b969862-6b09-4b63-8013-b639095d98ed_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p></p><div id="youtube2-J59CkteE9OM" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;J59CkteE9OM&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/J59CkteE9OM?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=J59CkteE9OM">YouTube</a> | <a href="https://open.spotify.com/episode/69MXlzFm4X6PPTSatZ0GkT">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/building-ai-products-for-the-other-95-brian-mcmullin/id1733103005?i=1000791103511">Apple</a></strong></em></p></div><p>Most of the AI products getting hype right now are built for people like us: the tech industry. But your customers don&#8217;t care how the AI works &#8212; they just want their problem solved in a way that&#8217;s easy for them.</p><p><a href="https://www.linkedin.com/in/brianemcmullin/">Brian McMullin</a>, SVP and Head of Product at Network Solutions, has spent 15 years building for small businesses at companies including HubSpot, ezCater, Wayfair, and SamCart. In that time, he&#8217;s yet to meet a business owner who wakes up thinking, &#8220;I can&#8217;t wait to use AI today.&#8221; For many of them, a prompt box might as well be a terminal window.</p><p>So his team doesn&#8217;t ship the box. They ship the finished work artifact, and all the user has to do is claim it.</p><p>In this episode, Brian shares:</p><ul><li><p>Why doing the work for small businesses first, instead of handing them an empty prompt box, is what gets them to try AI</p></li><li><p>His warning about baking inference into the core of your product before token pricing settles &#8212; and the two models he&#8217;s testing to find out what small businesses will actually tolerate</p></li><li><p>And what SamCart taught him about churn: customers left because basic things were broken, not because features were missing</p></li></ul><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. Nobody wakes up wanting to use AI</h2><p>Brian&#8217;s customers get maybe a couple of hours a week for anything that isn&#8217;t serving their own customers (marketing, website updates, invoicing, etc.). Nobody in that position is going to spend it learning what a good prompt looks like. So his team skips the box, and even the menu of agents to pick from:</p><blockquote><p>&#8220;Rather than showing them the prompt box, or even, &#8216;Hey, we have these agents, come use a marketing agent&#8217; &#8212; we need to take one or two steps further and say, &#8216;We&#8217;ve used this agent to map out your next two weeks of social media posts. Go schedule them, and if you want us to keep doing this every two weeks, turn it on right here.&#8217; Come and claim it.&#8221;</p></blockquote><p>The customer&#8217;s only job is to approve something that already exists.</p><p><strong>Product takeaway:</strong> Make the action you want to be the easiest one available. For non-technical users, that means doing the work speculatively and asking for confirmation, not instruction.</p><div><hr></div><h2>2. Proactive AI only works if it&#8217;s specific to that customer&#8217;s business</h2><p>There&#8217;s an obvious failure mode here, and Brian names it first: if you do the work for someone and make it generic, then you&#8217;ve just built a spam machine.</p><blockquote><p>&#8220;It can fall flat if it doesn&#8217;t feel personal&#8230; The tricky part there is what you need to drive that level of personalization is ultimately <strong>context</strong>.&#8221;</p></blockquote><p>His argument is that this is where established software companies have a real edge. Everyone has the same models, but not everyone has years of history about how a particular business actually operates.</p><p><strong>Product takeaway:</strong> Take inventory of what you already know about each customer that a competitor would have to learn from scratch.</p><div><hr></div><h2>3. Be careful about building inference into the parts of your product people already pay for</h2><p>Serving one more customer used to cost almost nothing. Inference broke that &#8212; and telling a small business that the same outcome now costs more is a hard message no matter how you frame it. This leads to a caution you don&#8217;t often hear from someone shipping AI features:</p><blockquote><p>&#8220;It&#8217;s actually almost a little bit of a warning of how deeply to embed AI into your core product and value proposition. Until we get to more stability in token pricing, the more you bake it into the core of what you do, the more you&#8217;re going to have to figure out how to have a predictable cost structure for your customers.&#8221;</p></blockquote><p>Domain registration works fine without AI, so Network Solutions has some insulation. Products where AI sits inside the workflow people already bought have none.</p><p><strong>Product takeaway:</strong> &#8220;AI-native&#8221; is a positioning decision with a variable cost attached. Work out which parts of your product would still be worth the subscription with the inference switched off.</p><div><hr></div><h2>4. Nobody has landed on the right way to charge for AI yet</h2><p>Brian&#8217;s team surveyed customers on pricing and found no consensus at all, so they&#8217;re running both common models: tiered usage, which risks the problem of paying for credits you don&#8217;t use, and a flat monthly price per agent, which puts a paywall in front of something customers haven&#8217;t tried yet. Worth listening to his breakdown of the tradeoffs on each. But the sequencing argument underneath matters more than the choice:</p><blockquote><p>&#8220;The biggest risk you have is that nobody actually uses the product in the first place. There is a temptation to build complex monetization models and put in tons of limits and over-engineer that aspect before you&#8217;ve truly found product-market fit.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Strict limits imposed early mostly buy you a smaller sample of the usage data you need to price correctly. Let people use it, then decide what to charge.</p><div><hr></div><h2>5. What SamCart taught Brian about churn</h2><p>At SamCart, Brian&#8217;s team was building good monetization tools for creators, and marketing was excellent at selling them. Then they read the churn reasons:</p><blockquote><p>&#8220;None of it was about these new features not driving them revenue. It was about, &#8216;I can&#8217;t get this basic thing to work. I need to add my team and I can&#8217;t do that. I can&#8217;t connect my payment processor in the way that I want to, and I&#8217;m not able to take payments.&#8217;&#8221;</p></blockquote><p>He also found that customers often buy for optionality and never touch the feature they bought for, which makes the core experience, not the headline feature, the thing that keeps them.</p><p><strong>Product takeaway:</strong> Retention is a lagging indicator, so argue for core experience work on frequency and breadth of usage instead. In the episode, Brian walks through how he makes that case to a CFO asking for ROI on a login page.</p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=J59CkteE9OM"><span>00:00</span></a><span> Introduction<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=131s"><span>02:11</span></a><span> Brian's product journey from developer to SMB product leader<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=245s"><span>04:05</span></a><span> How small businesses went from distrusting AI to have AI FOMO<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=453s"><span>07:33</span></a><span> SMBs don't wake up excited to use AI, they want to close the deal<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=605s"><span>10:05</span></a><span> Stop showing the agent, ship the finished work<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=791s"><span>13:11</span></a><span> Inference costs break the old SaaS margin math<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=983s"><span>16:23</span></a><span> Testing free credits and tiered usage with small businesses<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=1090s"><span>18:10</span></a><span> What usage-based pricing does to ARR and predictability<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=1431s"><span>23:51</span></a><span> Churn is a pile of micro-annoyances, not one event<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=1589s"><span>26:29</span></a><span> What SamCart's churn data revealed about broken basics<br></span><a href="https://www.youtube.com/watch?v=J59CkteE9OM&amp;t=1818s"><span>30:18</span></a><span> Conclusion</span></p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/brianemcmullin/">Brian&#8217;s LinkedIn</a></p></li><li><p><a href="https://www.networksolutions.com/">Network Solutions</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[Leader Spotlight: Building trust into real-time payments, with Abhijeet Ranadive]]></title><description><![CDATA[Abhijeet (Abhi) Ranadive is Chief Product Officer at Routable, where he leads product strategy for the company&#8217;s payments experience, including its instant payouts capability.]]></description><link>https://stories.logrocket.com/p/leader-spotlight-abhijeet-ranadive</link><guid isPermaLink="false">https://stories.logrocket.com/p/leader-spotlight-abhijeet-ranadive</guid><dc:creator><![CDATA[Jessica Srinivas]]></dc:creator><pubDate>Tue, 22 Sep 2026 07:03:55 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!c0d6!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><em><span>Abhijeet (Abhi) Ranadive is Chief Product Officer at Routable, where he leads product strategy for the company&#8217;s payments experience, including its instant payouts capability. Routable is a payouts platform that gives enterprise companies end-to-end orchestration across approvals, compliance, and reconciliation, so they can automate and scale high-volume payouts globally via a developer-first API. He previously served as VP of Product for Galileo FinTech at SoFi, and held product leadership roles at Intuit, Braintree (a PayPal service), and PayPal, following earlier work in enterprise architecture and API management at Cisco.</span></em></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!c0d6!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!c0d6!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!c0d6!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!c0d6!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!c0d6!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!c0d6!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png" width="895" height="597" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:597,&quot;width&quot;:895,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1299446,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://stories.logrocket.com/i/216049363?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!c0d6!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!c0d6!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!c0d6!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!c0d6!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd99f0c16-15ea-481c-8d2b-a863eb926ecd_895x597.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em><span>In this conversation, Abhi talks about what it takes to build instant payments into the core of a customer experience - from identifying the specific moments when speed actually matters to a customer, to keeping the underlying complexity of real-time money movement invisible to the people who depend on it. He shares how Routable rethought its product experience around trigger moments and conversion, the metrics that became essential once decisions had to be made in near real time, and why trust, rather than speed alone, is what keeps a payments product credible.</span></em></p><div><hr></div><h2><span>Rethinking the experience for instant payments</span></h2><h3><span>Instant payments aren&#8217;t simply a faster version of traditional payments &#8212; they change the economics of a marketplace. What do product teams need to rethink once payment becomes part of the customer experience rather than just the first step in a workflow?</span></h3><p><span>Customer empathy is a foundational pillar for our product team, and it&#8217;s one of our company&#8217;s values. For us, it was: How do we engage with empathy with our customers? We are a two-sided platform in many ways: the businesses sending money and the payees receiving it. What we uncovered was that they want to shorten the loop of earning and getting paid and the instant rails are increasingly in demand to accomplish that quickly.</span></p><p><span>When we talked to our customers, we found that they were very interested as well, because for them it was a retention lever. It was like, &#8220;Hey, my payees and stakeholders need it, and it will help us make them more satisfied and help them close that gap between earning and paying.&#8221; We needed to rethink our experience because it was not built for real time payments, and we had to find where the trigger moments were.</span></p><p><span>There are a lot of payees who are like, &#8220;Give me the money, I am fine whenever it comes.&#8221; but there are others who are really interested in instant payments. We wanted to capture the group that was super interested and had high intent. The other interesting insight was that there was a segment that was very interested in instant payments on the weekends &#8212; maybe they needed the money for a trip and couldn&#8217;t wait for Monday. So we realized that there was a group that always wanted instant and others that wanted it situationally at the right moment, but how do we find those moments and trigger the experience at that moment?</span></p><p><span>We had to think about the customer journey and figure out how to insert the experience, and what kind of experience could help make that choice. It&#8217;s almost like a checkout conversion. Once we have that buyer intent, how do we convert that intent into action as fast as possible and close that gap? Those were the two big aspects of rethinking the experience for instant payments.</span></p><p><span>There has been a shift in the market, too. There was a time when instant payments didn&#8217;t have much adoption, there wasn&#8217;t as much customer propensity to take it on. But with the new rails coming in &#8212; RTP and FedNow, and even stablecoins, we&#8217;ve seen a shift in the market in the last 12 to 18 months where instant has become the primary choice for all parties. In the past, the recipient always wanted their funds fast, and that makes sense. But now every party sees the benefit of having instant visibility into transactions. Instant has become a core pillar of our product strategy. We pride ourselves on paying anyone anywhere, including global markets, so having instant payment available not only in the U.S. but in international markets is the core strategy for our business.</span></p><h2><span>What customers actually wanted</span></h2><h3><span>Besides the segment that wanted payments faster on weekends, were there other customer needs you uncovered?</span></h3><p><span>From a payee standpoint, those were the two critical insights: There are some groups that say, &#8220;I always want it fast,&#8221; and the other group was a bit more situational.</span></p><p><span>One of the things we learned from the businesses was that, for them, instant was both a retention lever and a satisfaction lever. We also saw that instant could be a monetization lever, giving businesses an opportunity to turn payments into a profit center rather than simply a cost of doing business. This is what we are championing with our clients: If you can enable instant and close the gap between earn and pay, then you can drive satisfaction and retention for your payees and also create value for the business.</span></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive new posts every week.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h2><span>Balancing speed and trust</span></h2><h3><span>FinTech products require balancing speed, risk, and trust. Have there been situations where making a product faster would undermine customer trust, or the other way around?</span></h3><p><span>Trust is our real product. In payments, the businesses we serve value that they can trust us to send the money to the counterparty. The receiving parties trust that we will deliver the money to them. We always strive not to undermine this trust at any point, and we also have to ensure there is no operational complexity.</span></p><p><span>One of the things we&#8217;ve enabled well is what I call a product operations function, which is to look at the product and the needs of the clients post-launch of the product in the market. Many product teams are focused on building the product and launching. But I think the hard work comes after the launch, because that&#8217;s when you have to fine-tune the product and make sure it&#8217;s operationally viable. We talk about value and viability, and that is what PMs own. Most of the time viability is business viability, and that happens upfront. But if the operational viability for the product is lacking after the launch, it will be very hard to keep the product maintained in the market, no matter how valuable the product is. So we reversed it, and took a pre-mortem approach: What will go wrong, and what do we need to have upfront to make it operationally viable?</span></p><p><span>We spent a reasonably good amount of time on beta and pilot before going to GA. That wasn&#8217;t because we weren&#8217;t feature complete, we were feature complete in a traditional sense, but we wanted to build hardened requirements for supportability and operations. So even if one transaction doesn&#8217;t get to the intended party, or it&#8217;s late or delayed, we can have a remediation with the right tools. It&#8217;s all about visibility, having the observability, and then having a remediation action. I have a product operations lead on my team who was super focused on building those requirements upfront in the product build cycle, rather than after the fact.</span></p><h2><span>Leading the cross-functional build</span></h2><h3><span>Building products like instant payments requires close alignment across product, engineering, banking partners, fraud, operations, and customer support. What leadership practices become most important for that kind of cross-functional alignment?</span></h3><p><span>One of our key values is &#8220;Collaborate to win.&#8221; I strongly believe that successful teams have a clear north star metric that everyone is aligned to. It&#8217;s about shared outcomes and shared context. It&#8217;s easy to say this, but it&#8217;s hard to practice. Getting buy-in and alignment is where the real hard work is &#8212; communicating it out and getting the buy-in. Once you establish and get that buy-in, execution is smooth.</span></p><p><span>It takes a lot of upfront hard work to get that shared outcome and shared context. But, the hardest part was getting that alignment, not just with internal parties but with some of the external partners too. Once we got external partners to buy into the north star and gained that alignment, the rest of execution was much smoother.</span></p><h3><span>That must be difficult, too, since regulatory and compliance factors change over time.</span></h3><p><span>Compliance and risk are big for us, so we had those functions participating right from the outset. In fintech, regulatory compliance is part and parcel of the product design. The bigger issue to tackle was focus and prioritization. We took a few cycles to build it, considering the complexity of the product buildout. The real question was how do we keep teams focused and filter out all the noise? This is where a lot of the product leadership role comes in: Make sure your teams are focused, do continuous and rigorous prioritization, and make choices of what not to do, especially saying no to a lot of non-critical product requests, so the team can remain focused on what they need to deliver.</span></p><h3><span>Several of your products, including this one, required building entirely new capabilities rather than extending something that already existed. When you&#8217;re creating a new team around a technically complex problem like that, what do you focus on first?</span></h3><p><span>I&#8217;m a big fan of product vision north star, aligning the shared outcome and then sharing the context across all teams. That&#8217;s a crucial step, especially for a large cross-functional team. The first step is team formation, staffing the right people, this product required a village to build. But the other important part I want to call out is the right process to complement the team &#8212; we use this execution framework called Shape Up. It operates on a six-week cycle, and every six weeks you build something that is either demoable or releasable. It&#8217;s very different from agile methodology &#8212; it&#8217;s based on agile principles, but it takes it to the next level.</span></p><p><span>I want to tie this back to the product operating model: How do you create a truly empowered team? We&#8217;ve been working for a while on this concept of empowered teams of engineering, product, and design, that can take a problem, own it, work through the Shape Up methodology to break it down into six-week cycles with clear goals, and then execute on that. Creating the shared context and shared outcomes and getting that alignment upfront, plus the model of empowered teams using the Shape Up framework, has helped us move with high velocity toward the desired outcome.</span></p><h2><span>Data at the speed of instant</span></h2><h3><span>Instant payments don&#8217;t just speed up access to money &#8212; they also accelerate learning about customer behavior, like the insight about weekend timing. How did having those near real-time signals change the way you make product decisions?</span></h3><p><span>With instant payments, you have real-time data. Data-driven is one of the key principles we had, but instant took it to the next level. In a more traditional sense, you could probably get away with looking at data weekly, but in an instant scenario, we had to build metrics and dashboards so that almost at a daily level we are looking at the data and how things are performing. For example, not everyone is eligible for these instant offers, so we want to make sure that we continuously monitor the eligibility and how the funnel is converting.</span></p><p><span>We get real-time data and make decisions based on it. Our operating model changed to a consumer-centric mode where the data and decisions are real-time. Instant changes that paradigm for B2B. Now you&#8217;re in a consumer-centered model where you&#8217;re reviewing data in real time, looking at product behaviors and user behaviors, and adjusting levers in real time. Using those levers so we can adjust them in real-time to hit our outcome metrics was a substantial change for us.</span></p><h3><span>You&#8217;re analyzing data more frequently. Are there certain metrics that have become more valuable given the shorter payment cycle?</span></h3><p><span>At an overall level, the north star metrics didn&#8217;t change. But at a more granular level, we had to enable new metrics. One of which is the eligibility percentage: How do we maximize the percentage of transactions that qualify for instant?</span></p><p><span>This goes back to the tradeoffs we discussed on risk and compliance complexity, which are being managed through the eligibility criteria. We had to balance maximizing reach while minimizing compliance risk and fraud risk. We had to figure out how to balance them to have high eligibility and low risk.</span></p><p><span>The second lever was conversion. You may have a high intent, but the intent may not translate into completed action. So how do you close that gap between intent and action? We built an end-to-end metrics funnel, from notification email sent, to email opened, to button clicked, and then to accepting the instant offer. Enabling that funnel conversion metric was a big part of what we accomplished, and we&#8217;ve built a culture to make decisions based on this metric.</span></p><h2><span>Owning the complexity</span></h2><h3><span>Reducing the time between work and payment is clearly a benefit, but it adds complexity on the technical and operational side. How do you decide what parts of that complexity customers should experience, and which should remain invisible?</span></h3><p><span>One of our core product tenets is to own the complexity, it&#8217;s also Routable&#8217;s brand promise. We own the complexity so that there is no operational chaos on the customer side. We need to tackle regulatory complexity, fraud management complexity, and experience complexity. So, how do we make all of this simple and almost invisible to both parties, especially to the clients who are sending money? That was the key for us: building an experience that does not require much incremental effort for either party.</span></p><p><span>We inserted a trigger moment in the payment notification. If you&#8217;re already getting these notifications, that&#8217;s the right moment to make a choice about changing the payment speed. We had to figure out how to insert an experience that is seamlessly integrated into that notification trigger and doesn&#8217;t come across as a marketing offer. That was a core tenet for the experience design.</span></p><p><span>There was another level of complexity &#8212; technical architecture complexity. It&#8217;s not just about the payment; with the rails and the experience being instant, and all these real-time signals, the platform had to evolve to support instant money movement. Money movement is now happening in seconds versus minutes, so we decided to build a new platform to support the needs of the experience. That was another complexity we had to abstract away from both parties, we rolled it underneath the experience so it&#8217;s invisible to the customers.</span></p><p><span>You ask about tradeoffs. For Routable, it&#8217;s not a tradeoff of convenience and complexity. It&#8217;s really about trust - that&#8217;s our product. Our clients trust us to pay what is due, and our payees trust us that they&#8217;ll get what they are owed. That was something we didn&#8217;t compromise on. We won&#8217;t make tradeoffs on trust or on compliance and risk, but we will make tradeoffs on our timeline to get us to our goals.</span></p><h3>What does LogRocket do?</h3><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/?substack">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[Leader Spotlight: Building a platform-as-product culture, with Suneth Rupasinghe]]></title><description><![CDATA[Suneth Rupasinghe is Vice President of Global Enterprise Platforms & Services at HP, where he leads product management, engineering, and operations across SAP S4/HANA, ServiceNow, MS Dynamics, Adobe Commerce, MuleSoft, and HP&#8217;s global payment platforms &#8212; a portfolio supporting more than $50+ billion in annual revenue.]]></description><link>https://stories.logrocket.com/p/leader-spotlight-suneth-rupasinghe</link><guid isPermaLink="false">https://stories.logrocket.com/p/leader-spotlight-suneth-rupasinghe</guid><dc:creator><![CDATA[Katie Schickel]]></dc:creator><pubDate>Thu, 17 Sep 2026 07:02:48 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!_ol1!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><em><span>Suneth Rupasinghe is Vice President of Global Enterprise Platforms &amp; Services at HP, where he leads product management, engineering, and operations across SAP S4/HANA, ServiceNow, MS Dynamics, Adobe Commerce, MuleSoft, and HP&#8217;s global payment platforms &#8212; a portfolio supporting more than $50+ billion in annual revenue. He joined HP in 2022 as Vice President of B2C E-Commerce Solutions and Digital Solutions Architecture before moving into his current role in January 2025. Earlier in his career, he held engineering and architecture leadership roles at MetLife and Amplify Education, working across digital benefits platforms, e-commerce, and cloud-native modernization.</span></em></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!_ol1!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!_ol1!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!_ol1!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!_ol1!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!_ol1!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!_ol1!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png" width="895" height="597" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:597,&quot;width&quot;:895,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1316681,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://stories.logrocket.com/i/215691159?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!_ol1!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!_ol1!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!_ol1!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!_ol1!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c091dd3-a514-4995-bf6d-dc0dc0d3d44e_895x597.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p><em><span>In this conversation, Suneth talks about the three-year effort to transform HP&#8217;s enterprise platforms from maintenance-mode technology assets into true products &#8212; treating internal business units as customers, building a product management practice from scratch, and even reframing compliance itself as a product with its own roadmap. He also shares how that shift helped HP advance its compliance strengths, and how his team decides where AI capabilities should live as build-versus-buy questions get harder to answer.</span></em></p><div><hr></div><h2><span>From technology asset to product</span></h2><h3><span>What changed at HP that made the traditional approach to managing enterprise platforms no longer sufficient? What&#8217;s the reasoning behind the move to platform-as-product?</span></h3><p><span>Before taking on this role, I had spent much of my career in product development and bespoke software delivery, so I already approached technology through a product lens. When our platform organization came together about two years ago, it quickly became apparent that we were largely managing our platforms as technology assets rather than strategic products.</span></p><p><span>The conversations were typically centered on operational metrics: uptime, cost to operate, incidents, and service stability. Those are important measures, but they don&#8217;t tell you whether a platform is creating business value, enabling growth, or helping the company move faster. In many ways, it was a model I&#8217;d seen in other organizations as well.</span></p><p><span>At the same time, the pace of innovation across the industry was accelerating. Whether it was AI, automation, analytics, or new platform capabilities, vendors were delivering meaningful advancements at a rapid rate. Yet we often found ourselves unable to take advantage of them because of technical debt, heavy customization, or being multiple releases behind. Our business partners would increasingly ask, &#8220;If the platform can do this, why can&#8217;t we?&#8221; That question became an important catalyst for change.</span></p><p><span>We also recognized that our investment approach was too focused on maintenance and sustainment. What was missing was a strategic roadmap for how these platforms should evolve over time. Reusability, adoption, modernization, and innovation were not being managed as intentional outcomes. As a result, teams frequently had to build capabilities from scratch, increasing both cost and time to value. The move to platform-as-product addressed those challenges. It shifted the conversation from maintaining technology to continuously evolving capabilities that serve the business. Each platform now has a clear strategy, a roadmap, defined outcomes, and measurable value drivers, just as any product would.</span></p><p><span>When we looked at all of those factors together, it became clear that a different operating model was required. Treating platforms as products gave us a framework to modernize more effectively, accelerate adoption of new capabilities, improve reuse, and align platform investments more directly with business outcomes. That&#8217;s ultimately why we came together as a team and why the platform-as-product model became such a central part of our strategy.</span></p><h2><span>Building the product management practice</span></h2><h3><span>Can you briefly explain who your customers are on these platforms?</span></h3><p><span>We&#8217;re fundamentally an IT-for-IT organization. Our primary customers are our peers across the CIO organization, particularly the value-aligned teams responsible for delivering business outcomes. These teams work closely with functions such as sales, supply chain and finance, translating business objectives into technology-enabled capabilities.</span></p><p><span>Our role is to provide the platform foundation that allows those teams to move quickly and effectively. We focus on ensuring the right capabilities, services, and products are available so they can build and deliver solutions that drive the outcomes their business partners are seeking.</span></p><p><span>In that sense, our direct customers are the value-aligned teams themselves. By enabling them with scalable, reliable, and continuously evolving platform capabilities, we help accelerate delivery and reduce the complexity of building solutions from scratch.</span></p><p><span>Beyond that, there is a natural ripple effect. The value-aligned teams serve the business, and the business serves HP&#8217;s customers. So while our day-to-day engagement is primarily with our peers inside the CIO organization, the capabilities we deliver ultimately contribute to business performance and, in some cases, directly influence the experience of HP&#8217;s end customers.</span></p><p><span>That&#8217;s why we think of our platforms not simply as technology assets, but as business enablers. Our success is measured by how effectively we help our internal customers achieve their outcomes, which in turn supports broader business and customer success.</span></p><h3><span>So how did you start building the product management practice?</span></h3><p><span>When we brought the organization together, one thing was clear: we needed to evolve how we operated. We had incredibly strong platform talent and deep domain expertise, but product management required a different set of capabilities. The challenge wasn&#8217;t the quality of the people; it was creating the structure, skills, and mindset needed to operate as a product organization.</span></p><p><span>We focused on two priorities. First, we brought in the right leaders to establish and scale a modern product management practice. Second, we built an operating model to support it. That operating model became the foundation for how we worked, how we developed talent, and how we drove consistency across the organization.</span></p><p><span>A critical part of the journey was investing in the people already here. We wanted to build capabilities from within, not simply hire them from outside. Relying exclusively on external talent creates unnecessary friction and misses the opportunity to leverage the deep institutional knowledge that already exists. As a result, training, coaching, and upskilling became core elements of the transformation.</span></p><p><span>That investment continues today. We have ongoing learning programs for product managers, engineers, and platform teams, and we work closely with strategic partners to stay current on emerging practices and technologies. More recently, that focus has expanded to include AI-native product management and how AI can enhance product strategy, delivery, and customer engagement.</span></p><p><span>Ultimately, our approach has remained consistent: bring in new talent and perspectives where needed, while continuously developing the people already in the organization. The operating model ties those two elements together and helps create a sustainable product management community that continues to grow, adapt, and mature over time.</span></p><h3><span>What do you look for now when you&#8217;re hiring product managers, in terms of AI skills?</span></h3><p><span>This has evolved significantly over the last 18 months. The foundational capabilities of product management are still essential. We look for people who have successfully managed products, developed strategy, worked closely with customers, navigated complex stakeholder environments, and consistently delivered business value. Those fundamentals haven&#8217;t changed.</span></p><p><span>What has changed is the expectation around AI. We now look for product managers who understand how to leverage AI-enabled tools to amplify their effectiveness. It&#8217;s not just about working faster; it&#8217;s about making better decisions, conducting deeper research, uncovering insights more quickly, and improving the overall quality of outcomes. Product managers today have access to capabilities that simply didn&#8217;t exist a few years ago, and we&#8217;re looking for people who know how to incorporate those tools into their day-to-day work.</span></p><p><span>The second area we&#8217;re increasingly focused on is what we call the full-stack builder. This is an evolution of the traditional product manager role. In addition to owning product strategy, customer engagement, and roadmap execution, these individuals have a strong understanding of engineering practices, the software development lifecycle, and agentic AI systems. They know how to orchestrate work across people, platforms, and AI agents to accelerate delivery and innovation.</span></p><p><span>We&#8217;re actively bringing this capability into the organization today. While traditional product management skills remain critical, we believe AI-native product managers and full-stack builders will become two of the most important roles over the next 18 to 24 months as we continue to transform.</span></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive new posts every week.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h2><span>Learning to listen to internal customers</span></h2><h3><span>When the customers for your products are your internal business teams, how do you understand what they need and translate that into a product roadmap?</span></h3><p><span>Serving internal customers is different because a lack of feedback isn&#8217;t necessarily a positive signal. If internal teams aren&#8217;t engaging with your platform, they&#8217;ll often find alternative ways to solve the problem. In many cases, silence is actually a sign that you&#8217;re not creating enough value.</span></p><p><span>Early on, we recognized that adoption couldn&#8217;t be a pull model. It had to start as a push. Our responsibility was to proactively engage with customers, understand their challenges, and demonstrate how the platform could help them achieve their goals. As the platform matured and delivered value, that dynamic gradually shifted from push to pull.</span></p><p><span>We borrowed a concept from leading product organizations: Customer Success Management. We asked our product managers to act as customer success leaders, not just product managers. A key part of their  role was building relationships with internal customers, providing white-glove support, and helping teams realize value from the platform. We held workshops, monthly ideation sessions, and regular business engagements focused on understanding objectives, exploring use cases, and identifying opportunities for rapid delivery.</span></p><p><span>One of the most important mindset shifts was moving away from a requirements-first approach. Instead of starting with everything a customer wanted, we started with what the platform could deliver today to help them achieve business outcomes faster. From there, we could iterate and expand capabilities over time. That approach significantly accelerated time to value.</span></p><p><span>We were also applying what is now commonly referred to as forward-deployed engineering long before the term became popular. We would rapidly build a proof of concept, validate the use case, and demonstrate the outcome. That hands-on engagement created momentum, accelerated adoption, and strengthened the partnership between our platform teams and the value aligned teams.</span></p><p><span>Ultimately, the transition from push to pull came from one principle: ensuring our customers were successful. When customers experience value quickly, adoption follows naturally, and the relationship evolves.</span></p><h2><span>One strategy, five platforms</span></h2><h3><span>How do you create a consistent product strategy across platforms as different as SAP, ServiceNow, payments, commerce, and integration?</span></h3><p><span>The goal isn&#8217;t to make every platform do the same thing. These platforms serve very different purposes. Adobe Commerce is customer-facing, SAP is deeply domain-specific, and MuleSoft is an integration platform that enables connectivity across the enterprise. A one-size-fits-all strategy simply doesn&#8217;t work.</span></p><p><span>What you can do is create a common strategic framework that applies across all platforms while allowing for platform-specific execution. We developed a strategy built around five pillars: simplification, adoption and maturity, innovation ignition, operational excellence, and compliance and security. Those pillars are intentionally platform-agnostic. They provide a consistent way to think about progress and investment without prescribing the same outcomes everywhere.</span></p><p><span>For example, simplification means something very different on SAP than it does on MuleSoft. In SAP, it&#8217;s about reducing customization and moving toward a clean core so upgrades can happen faster and with less effort. In MuleSoft, simplification is about increasing reuse and standardization so teams can leverage existing APIs and integrations instead of continually building new ones. The outcomes are different, but both support the same strategic objective.</span></p><p><span>Our platform product managers helped shape the framework from the beginning. We came together to define the overarching strategy and then built FY26 roadmaps aligned to each of the five pillars. The initiatives within those roadmaps vary by platform, but they all connect back to the same strategic themes. That&#8217;s how we standardize. It gives you the framework to talk about in a consistent way and the flexibility to be specific where it matters.</span></p><h3><span>With competing demands across so many platforms, what gets prioritized?</span></h3><p><span>Once we established the framework, we assessed every platform against a maturity model across each strategic pillar. The model included roughly fifteen subcategories and measured capabilities on a spectrum from reactive to leading. That assessment gave us an objective view of where we stood and created a heat map that clearly highlighted areas of strength and areas requiring attention.</span></p><p><span>From there, prioritization became much more straightforward. We always have a finite investment envelope, so we direct funding toward the highest-value opportunities where maturity is lowest. The goal is to elevate those critical capabilities while sustaining the areas that are already performing well.</span></p><p><span>When we built our FY26 roadmap, we used a simple two-dimensional lens: business value and maturity. We focused our investments on the capabilities that would deliver the greatest impact while moving the most significant gaps forward. In practical terms, that meant turning the red areas into yellow, while ensuring the green areas remained strong. It&#8217;s deliberately simple, and it&#8217;s the approach that&#8217;s held up for us.</span></p><h2><span>Treating compliance as a product</span></h2><h3><span>How do you build compliance into a platform in a way that lets teams move faster rather than slowing them down?</span></h3><p><span>Traditionally, compliance happens at the end of the process. Teams build and deliver, and then they encounter a series of control gates that require rework, delays, and additional effort. We wanted to change that dynamic.</span></p><p><span>Compliance was one of the five pillars of our strategy because we had clear visibility into where the gaps and friction points existed across our platforms. Rather than treating compliance as a set of checkpoints, we treated it as a product.</span></p><p><span>When people hear &#8220;product,&#8221; they often think of software, but in this case it&#8217;s a service-oriented product. Take a common ITGC requirement like timely incident management. The product isn&#8217;t just a control. It&#8217;s the end-to-end capability: the process definition, stakeholder responsibilities, workflows, reporting, and services that enable customers to meet their compliance obligations with minimal effort.</span></p><p><span>The key shift was moving the focus from what we needed to do internally to what our customers needed to be successful. We built a service layer around compliance requirements and packaged those capabilities into reusable products that teams could consume consistently.</span></p><p><span>We&#8217;ve also matured those products over time. In the early stages, many of these services were highly manual. We then introduced automation and standardized reporting, creating semi-automated capabilities. Today, some of those same products have evolved into AIOps-based services supported by autonomous agents. The underlying objective hasn&#8217;t changed, but the delivery model has become increasingly self-service, intelligent, and scalable.</span></p><p><span>That&#8217;s how compliance becomes an accelerator rather than a constraint. When you build it into the platform as a product, teams no longer encounter compliance at the end of the journey. They consume it throughout the process as a built-in capability, allowing them to move faster while strengthening governance and control.</span></p><h3><span>You strengthened the organization&#8217;s compliance posture. What organizationally and technically contributed to that result?</span></h3><p><span>What made the difference was treating compliance as a product rather than a project. Too often, organizations overly focus on meeting only the immediate needs. We took a different path and approached compliance as a transformation. That meant redefining processes, establishing clear service ownership, improving repeatable compliance products and controls, and helping teams build on their understanding that compliance is a core business capability. The goal was to more deeply embed compliance into the way we operate.</span></p><p><span>It wasn&#8217;t a single technical change or organizational initiative. It required aligning people, processes, and technology around a common operating model and strengthening a structure that scales consistently across the organization. This demands tremendous commitment from the team. But that investment leads to a stronger foundation. Today, we have further developed processes, reusable compliance products, and greater clarity around expectations and accountability. As a result, we&#8217;re able to move faster, scale more effectively, and continuously strengthen our compliance posture.</span></p><p><span>The key to this mindset is recognizing that this is never about solving isolated compliance requirements. It is about continuing to strengthen a capability that improves how the organization operates year after year.</span></p><h2><span>What comes next: Signals, and the AI build-versus-buy question</span></h2><h3><span>What signals tell you that a platform is becoming a successful product rather than simply a well-run technology?</span></h3><p><span>The strongest signal is when customers start coming to you asking how they can build on the platform, what capabilities already exist, and what they can reuse. That&#8217;s the point where the conversation shifts from operating technology to consuming a product. Instead of us driving adoption, customers are actively seeking ways to leverage the platform to achieve their outcomes.</span></p><p><span>We also track a set of metrics to understand whether we&#8217;re moving in the right direction. Platform Net Promoter Score (NPS) gives us insight into customer sentiment, while measures such as reusability, maturity, and cost per transaction help us evaluate how effectively the platform is delivering value. As we execute against our roadmaps, we expect those maturity scores to increase over time, providing a quantitative view of progress.</span></p><p><span>For me, though, the most meaningful indicator is the transition from push to pull. In the early stages, we had to invest heavily in customer engagement, advocacy, and adoption. We were constantly demonstrating value and encouraging teams to take advantage of platform capabilities. Today, on some of our platforms, conversations have fundamentally changed. Customers are approaching us with questions such as, &#8220;Can I use this capability?&#8221; or &#8220;When will this feature be available?&#8221; Many of those discussions are centered on AI and automation, where demand is often outpacing our ability to onboard teams responsibly.</span></p><p><span>We&#8217;re not at that point across every platform yet, and there&#8217;s still work to do. But where we are seeing strong customer pull, increasing reuse, improving maturity metrics, and growing demand for new capabilities, those are clear signs that the platform is evolving into a product rather than simply being managed as a piece of technology.</span></p><h3><span>With AI capabilities coming from enterprise platforms, external vendors, and your own teams, how do you decide what to build, what to buy, and where each capability should live?</span></h3><p><span>We start with our broader enterprise AI strategy, which is led by our Applied AI team. We complement that with a platform-native AI strategy supported by a set of guiding principles that help us make decisions in a rapidly changing landscape.One of the most important principles is augmentation versus differentiation. We ask a simple question: is this capability creating meaningful competitive differentiation, or is it augmenting an existing process? If it&#8217;s primarily augmentation, our default position is to leverage native platform capabilities. It&#8217;s typically easier to implement, easier to maintain, and allows us to benefit from ongoing innovation within the platform ecosystem.</span></p><p><span>We also evaluate total cost of ownership. We&#8217;ve already made significant investments in our core platforms and licensing. Wherever possible, we want to maximize the value of those investments before introducing additional technologies that increase complexity and support costs.</span></p><p><span>Another key consideration is technology sprawl. As AI lowers the barriers to building and acquiring solutions, organizations risk creating fragmented technology landscapes. Like many large enterprises, we&#8217;re actively working to simplify and consolidate. We want innovation, but we want it in a way that&#8217;s intentional and sustainable.</span></p><p><span>That said, if a capability is truly differentiating and can accelerate business outcomes, we&#8217;re willing to bring in external solutions or build capabilities ourselves. The important thing is making that decision deliberately. What may be differentiating today can become a standard platform capability tomorrow. That&#8217;s why strong partnerships with platform vendors are so important. We invest heavily in those relationships to gain early visibility into their roadmaps, participate in preview programs, and influence future capabilities. In some cases, we may choose a third-party solution knowing that a comparable native capability is likely to become available in the near future. When that happens, we can transition to the platform-based solution through a planned and intentional lifecycle.</span></p><p><span>We&#8217;ve deliberately avoided creating a rigid decision matrix because the market is evolving too quickly for static rules to remain effective. Instead, we apply a consistent set of principles around differentiation, augmentation, total cost of ownership, platform leverage, and technology sprawl. Those principles don&#8217;t guarantee every decision will be perfect, but they ensure each decision is thoughtful, transparent, and aligned with our long-term platform strategy.</span></p><h3>What does LogRocket do?</h3><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/?substack">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[What Do You Do When AI Agents Replace Your Human Users? | Brandon Harris, VP of Product at Pantheon]]></title><description><![CDATA[Pantheon's VP of Product, Brandon Harris, on what happens to your business when most of your traffic stops being human, and why the web becoming one giant CMS makes brand and trust more valuable.]]></description><link>https://stories.logrocket.com/p/what-do-you-do-when-ai-agents-replace-your-human-users-brandon-harris</link><guid isPermaLink="false">https://stories.logrocket.com/p/what-do-you-do-when-ai-agents-replace-your-human-users-brandon-harris</guid><dc:creator><![CDATA[Jeff Wharton]]></dc:creator><pubDate>Tue, 15 Sep 2026 12:57:20 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/5e0a95d8-a595-4e12-8c81-e09652f46164_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div id="youtube2-1frEz9zQvMw" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;1frEz9zQvMw&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/1frEz9zQvMw?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=1frEz9zQvMw">YouTube</a> | <a href="https://open.spotify.com/episode/6iJkfHKMKBHZPGZudxHrZa">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/what-do-you-do-when-ai-agents-replace-your-human-users/id1733103005?i=1000789780639">Apple</a></strong></em></p></div><p><span>Every day, more of your traffic isn&#8217;t human. AI agents show up, grab what they need, and leave &#8212; and </span><em><span>you</span></em><span> pay to serve every visitor, whether they see your brand or not.</span></p><p><span>Today&#8217;s guest has seen this from both sides. </span><a href="https://www.linkedin.com/in/cxoplus/"><span>Brandon Harris</span></a><span> is VP of Product at Pantheon, the WebOps platform behind a huge slice of the WordPress and Drupal world. But before that, he was sending the bots &#8212; at Wiser, where his team scraped millions of pages a day to power retail pricing intelligence.</span></p><p><span>His take: agents aren&#8217;t killing the web. They&#8217;re just the new way information is traveling. And the companies that adapt are about to pull way ahead of the ones that don&#8217;t.</span></p><p><span>In this episode, Brandon shares:</span></p><ul><li><p><span>How AI is turning the whole web into one giant CMS and quietly rewriting the job of anyone who publishes content</span></p></li><li><p><span>Why site visits are the new impressions, and what happens to hosting economics when the meter can&#8217;t differentiate a customer from an agent</span></p></li><li><p><span>Why trying to keep agents out is the modern-day version of blocking Google from indexing your site, and what to do instead</span></p></li></ul><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. Agents aren't killing the web. They're just the next mode of transportation.</h2><p>The &#8220;dead internet&#8221; theory assumes that this new wave of internet traffic is essentially eating itself; AI made to consume AI and AI made to be consumed by AI. Brandon has a different opinion, and he relates it to the history of transportation:</p><blockquote><p><span>&#8220;We were walking, and then we found wheels, and carts, and animals, and then bicycles. We could do some self-powered things, and then trains, then cars. But each time that happens, there&#8217;s this sense that the new technology for transportation is gonna overtake all of the previous technology, and we&#8217;ll never use it. But the reality is, I still have to teach my kids how to ride a bike.&#8221;</span></p></blockquote><p><span>If you replace the thing being transported from people to information, then agents are a genuinely new transportation method. As a result, browsing traffic is changing. But it&#8217;s not being eliminated.</span></p><blockquote><p><span>&#8220;We&#8217;re gonna see a compression in the volume of human traffic, and we&#8217;re gonna see a compression in the volume of traffic that is intended just for browsing, but it&#8217;s never gonna go away. And in some ways, it makes it more valuable.&#8221;</span></p></blockquote><p><strong><span>Product takeaway:</span></strong><span> Falling human traffic isn&#8217;t an automatic sign of failure. The mix is changing underneath the number. As a result, you should </span><strong><span>segment your analytics by traffic type</span></strong><span> before you react to the topline. Because the humans who still show up are arriving with </span><strong><span>more intent</span></strong><span> than they used to, and that should change what you build for them.</span></p><div><hr></div><h2>2. Why site visits are the new impressions</h2><p><span>Here&#8217;s the uncomfortable math: a bot and a buyer cost your CDN, host, and CMS the same amount to serve. </span></p><p><span>Brandon, who spent a stint as interim CMO at a previous company, sees the same correction coming that advertising already went through: going from counting eyeballs to paying for </span><strong><span>action</span></strong><span>.</span></p><blockquote><p><span>"The visit will become kind of like impressions. They're valuable, but not the same value as a conversion. And we'll start to get better at tracking both."</span></p></blockquote><p><span>The interesting part is that this isn&#8217;t a case for locking the gates, so to speak. Brandon&#8217;s point is that the industry posture is shifting from </span><em><span>keep everything out</span></em><span> to something more selective &#8212; closer to </span><strong><span>traffic grading</span></strong><span>, where a site decides what a given interaction is actually worth and both sides negotiate from there. Some large publishers are already building toward charging for access.</span></p><p><span>He relates it to sites in the past blocking Google from indexing them:</span></p><blockquote><p><span>&#8220;Five years ago, everything was: how do we block bots, how do we stop them from coming? And it&#8217;ll shift to, well, we want some, but not all [agents].&#8221;</span></p></blockquote><p><strong><span>Product takeaway:</span></strong><span> If your economics are priced per visit but your value is created per conversion, you have a modeling problem that&#8217;s going to get worse every quarter. </span><strong><span>Start instrumenting traffic by value</span></strong><span> now &#8212; which agents drive real conversions and which are pure cost &#8212; so that when better grading and pricing tools arrive (Brandon guesses roughly a year, and probably from startups rather than infrastructure incumbents), you already know what you&#8217;d pay for.</span></p><div><hr></div><h2>3. Is the whole web becoming one giant CMS?</h2><p>Traditionally, <span>a CMS holds structured information and assembles it into a page on request. Now, that&#8217;s a pretty fair description of what an agent does to the entire internet.</span></p><p><span>Brandon watched an early version of this at Gannett fifteen years ago, where a project assembled news sites on demand from whatever content was available across hundreds of properties. The difference now is that the source pool is everyone&#8217;s site, not just yours:</span></p><blockquote><p><span>&#8220;If the web is the CMS, if all this data is available very quickly in real time, I can assemble it into whatever I need it to be to either build a new product or a new page on demand.&#8221;</span></p></blockquote><p><span>He&#8217;s clear that this produces slop, too, not just value &#8212; the same split we get across all platforms. Which is why the job of publishing content changes. It&#8217;s no longer enough to just have accurate information sitting on your domain. Your content now has to survive extraction and still be recognized as </span><em><span>yours</span></em><span>.</span></p><blockquote><p><span>&#8220;How do you provide brand placement within the content that you create, so that when it is pulled, it is still representative of your brand or harkens back to your brand?&#8221;</span></p></blockquote><p><span>Brandon frames this as inserting your brand&#8217;s &#8220;fingerprints&#8221; into your content the way broadcast used product placement. The industry&#8217;s done this before: Google forced an entire generation of teams to learn how to structure content for a crawler. This is the same thing, just maybe one level up in complexity.</span></p><p><strong><span>Product takeaway:</span></strong><span> Audit your content for what survives being pulled out of context. Does a paragraph lifted into an answer still carry your point of view, data, name, etc.?</span></p><div><hr></div><h2>4.  Look for the drift before it wrecks the tool</h2><p><span>Unhappy users rarely file a ticket saying they&#8217;ve lost confidence in you or your product suite. They just start to drift, slowly, in a way that never shows up as an event.</span></p><p><span>Brandon&#8217;s framing for what to do about it comes in two parts:</span></p><p><span>First, know what you&#8217;re protecting:</span></p><blockquote><p><span>&#8220;Understand your superpower and your kryptonite. Your superpower, amplify it &#8212; get your super suit, get the utility belt. For your kryptonite, protect against it, because it&#8217;s not just things that you&#8217;re weak at, it could be things that drain you and drain your value.&#8221;</span></p></blockquote><p><span>Second, go hunting for change on purpose, because it won&#8217;t announce itself. He borrows the term from machining:</span></p><blockquote><p><span>&#8220;If you&#8217;re not looking for the drift, you won&#8217;t see it until it&#8217;s messed up your tool.&#8221;</span></p></blockquote><p><span>This means watching real user sessions, tracing where people fall out of the funnel and why, and treating behavioral shifts as information rather than noise. This loops straight back to agents: the mix of searchers and browsers on your site is changing right now. If AI handles the initial lookup, then maybe your search bar matters less, and your browsing experience matters more. You won&#8217;t know unless you go and check.</span></p><div><hr></div><h2><span>Chapters</span></h2><p><a href="https://www.youtube.com/watch?v=1frEz9zQvMw"><span>00:00</span></a><span> Introduction<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=144s"><span>02:24</span></a><span> Brandon's product journey: From film school to Wiser and Pantheon<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=226s"><span>03:46</span></a><span> The dead web debate: Agents are just the next mode of transport<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=433s"><span>07:13</span></a><span> Why traffic-based pricing can break when the majority of your visitors are agents<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=577s"><span>09:37</span></a><span> Site visits are becoming the new impressions<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=728s"><span>12:08</span></a><span> Grading traffic and the identity arms race it will start<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=933s"><span>15:33</span></a><span> When the whole web becomes your CMS<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=1082s"><span>18:02</span></a><span> Putting your brand's fingerprints in content that AI will pull without accessing your site<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=1365s"><span>22:45</span></a><span> How people decide to trust a site<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=1659s"><span>27:39</span></a><span> Watching sessions to catch user drift before customers quietly leave<br></span><a href="https://www.youtube.com/watch?v=1frEz9zQvMw&amp;t=1951s"><span>32:31</span></a><span> Conclusion</span></p><h3><span>Links</span></h3><ul><li><p><a href="https://www.linkedin.com/in/cxoplus/"><span>Brandon&#8217;s LinkedIn</span></a></p></li><li><p><a href="https://pantheon.io/"><span>Pantheon</span></a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/">LogRocket.com</a>.</p><p></p><p></p><p></p><p></p><p></p>]]></content:encoded></item><item><title><![CDATA[Leader Spotlight: Why inclusivity is a growth strategy, not a nice-to-have, with Jennifer Schwartz]]></title><description><![CDATA[Jennifer Schwartz is VP of Digital Product at Universal Audio, where she leads digital product strategy for the pro audio brand.]]></description><link>https://stories.logrocket.com/p/leader-spotlight-jennifer-schwartz</link><guid isPermaLink="false">https://stories.logrocket.com/p/leader-spotlight-jennifer-schwartz</guid><dc:creator><![CDATA[Katie Schickel]]></dc:creator><pubDate>Tue, 15 Sep 2026 07:02:49 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!T7we!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><em><span>Jennifer Schwartz is VP of Digital Product at Universal Audio, where she leads digital product strategy for the pro audio brand. Before that, she spent over six years as VP of Digital Product at Fender Musical Instruments Corporation, where she built the company&#8217;s entire consumer-facing digital ecosystem from the ground up &#8212; including Fender Play, a subscription learning platform with more than 2,000 video lessons, and Fender Tune, Fender&#8217;s first-ever mobile app. She has also held digital product and content leadership roles at The Beachbody Company, Live Nation Entertainment, and Walt Disney Internet Group, working across music, entertainment, and health and fitness.</span></em></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!T7we!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!T7we!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!T7we!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!T7we!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!T7we!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!T7we!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png" width="895" height="597" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:597,&quot;width&quot;:895,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1328263,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://stories.logrocket.com/i/215127153?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!T7we!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!T7we!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!T7we!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!T7we!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F836b8b8a-f050-458c-b829-63f8fbbfa426_895x597.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em><span>In this conversation, Schwartz talks about what inclusivity actually means for a product team &#8212; not a single interface that tries to serve everyone the same way, but a product that lets a first-time player and a 25-year veteran each find their own moment of joy. She discusses building teams that include both the expert and the beginner in the room, translating jargon like reverb and distortion into something a new player can feel rather than just read, and why she pitches inclusivity to leadership not as a values statement, but as a growth number.</span></em></p><div><hr></div><h2><span>Redefining who counts as a musician</span></h2><h3><span>Who has the musical instrument industry traditionally been designed and built for?</span></h3><p><span>It depends. Inclusivity can relate to any category. Even though I happen to be in the music industry, it comes down to who is your audience, who is the audience you want to grow into, and how can you include them? It can be anything or anyone.</span></p><p><span>There are some companies who target a user in a particular part of their journey. I&#8217;m going to use some of the more school-focused, education-based products that are targeting that younger demographic, newer to their journey. They&#8217;re specifically targeting that person, but they can grow as that person learns. They can continue to grow with that company.</span></p><p><span>But what I&#8217;m talking about in my experience is more the popular music, the recording-based music, or fretted instruments, as they say. Who is that customer? Who do they design for? I think largely it becomes and has become the player. They want to talk to, speak to, and engage with those who use their products. Usually, once you start playing &#8212; let&#8217;s use guitar for example &#8212; you start understanding those things within a year or so, and as you grow in your journey, you can continue that for a lifetime. You are a player, so therefore you fit into their category.</span></p><h3><span>Who&#8217;s making music now that wouldn&#8217;t have called themselves a musician 10 years ago, and how does that redefine the audience?</span></h3><p><span>You nailed it. That&#8217;s exactly why this becomes an interesting conversation now. Over the course of 10 years, we could boil it to five years, even two years, and the definition of who is a musician, who can create music, has dramatically changed. Going back to pre-AI, pre- a lot of the technology that has grown to support new players, that target audience for MI companies 10 years ago would have had to invest a good amount of time and money to be comfortable enough to be one of their engaged users.</span></p><p><span>But that changed&#8212;needle scratch&#8212;to a point where you no longer need years of formal musical training to begin creating music. All you need is to be able to access your keyboard and an app, and you are making music to a degree that you can be composing orchestras. You could be making beats, designing and writing rock bands, without ever having to know one note or be able to sing one note. That&#8217;s how it changed so dramatically and why it&#8217;s so interesting.</span></p><h2><span>Designing across the beginner-to-pro spectrum</span></h2><h3><span>How do you build a single product experience for someone playing their first chord and someone who&#8217;s been playing for 25 years?</span></h3><p><span>That&#8217;s where it gets a little complicated because, from my perspective, trying to make sure we are inclusive to all doesn&#8217;t mean you have to have a one-size-fits-all solution &#8212; one onboarding journey, one UI that has to satisfy somebody who just picked up an instrument for the first time versus someone who&#8217;s been playing for 25 years. It doesn&#8217;t have to be the same. You just need to make sure that each of those categories, in their extreme, can do what they want to do, can engage with the product, can find that moment of joy as easily as the other one. That&#8217;s all you have to do.</span></p><p><span>I&#8217;ve seen established music companies say, &#8220;We don&#8217;t want to look stupid, or talking down or condescending to a player who knows exactly what they&#8217;re doing. We don&#8217;t want to ask, &#8216;do you know how to play a chord?&#8217;&#8221; Of course you don&#8217;t, but two things: First, you need to support that user at every step of the journey. And you would be surprised how open more established players are to making sure other people can learn &#8212; they&#8217;re not going to turn their nose at why you&#8217;re offering this kind of information. And, quite frankly, we have sophisticated MarTech tools right now to make sure you don&#8217;t even have to see that information if it doesn&#8217;t pertain to you.</span></p><h3><span>If you don&#8217;t want to insult that power user, where&#8217;s the line between making something effortless and making the skill behind it feel unnecessary?</span></h3><p><span>That&#8217;s exactly the fine line that one has to traverse: to make sure you don&#8217;t condescend to a pro user, but that you&#8217;re also inclusive. That&#8217;s sort of where the experience starts, and it&#8217;s where, from a product perspective, I get really interested: How do you do that? What language do you use? What does it look like, and what does it feel like? We need to maintain the brand integrity &#8212; we don&#8217;t want to look like we&#8217;re pandering or going beyond who we are as a product. That&#8217;s the real fine line, but it&#8217;s really important.</span></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive new posts every week.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h2><span>Building inclusive teams to build inclusive products</span></h2><h3><span>Are there any techniques or frameworks you use &#8212; questions you constantly ask, or mantras you have?</span></h3><p><span>I do it rather intentionally, and how I do it is by creating the team itself to develop a feature. Say we&#8217;re going to develop a feature, an onboarding, or roll out a new product. I like everything to be collaborative, so it&#8217;s not just one vision or one focus. It&#8217;s been crucial to me to build my teams to make sure we have somebody to represent each side inherently. We have a player who really does know, speaks the language, and knows what they want and don&#8217;t want. Conversely, we need the person who is just beginning their journey. That&#8217;s the most important thing, having the two together. As we&#8217;re ideating, as we&#8217;re doing wireframes and user flows, both sides of that user are in mind, always, in the room at every step of the journey.</span></p><p><span>You&#8217;re designing something, and someone will use a word or a phrase that a beginner wouldn&#8217;t understand, and the beginner says, &#8220;Wait, what is that?&#8221; &#8220;Oh, yeah, you don&#8217;t know that yet&#8221; &#8212; so how do we support you at that moment? That&#8217;s how it&#8217;s building that team. It&#8217;s not even just about product people &#8212; it&#8217;s about your content providers, your designers, your UI, UX. Everything should in some way have that breadth and depth of experience in the room for collaboration.</span></p><h3><span>How does who&#8217;s actually in the room designing a product change what the product ends up saying to the people using it?</span></h3><p><span>Just data. We start with our existing data &#8212; what do we know about this person or this audience from a data perspective, and how can we apply that as sort of a shot in the dark, an assumption theory? Then we plot that out as quickly as possible, get to some MVP, or minimum viable product, and get that out there, and start testing and iterating: Were we right? Are we wrong? Should we adjust? Should we pull back? It really just starts with that best guess.</span></p><h2><span>Speaking to different fluency levels</span></h2><h3><span>When a technical term means everything to a veteran and nothing to a beginner, how do you decide whether to teach the word or design around it?</span></h3><p><span>That&#8217;s something I&#8217;m still experimenting with. That question is so relevant when you&#8217;re talking about music &#8212; how do you describe sound? How do you explain distortion or delay or reverb to somebody who has no idea? They&#8217;re words, and you could read the words, but until you actually hear it, you go, &#8220;Oh, that.&#8221; That&#8217;s a big challenge. But as we move toward these more experiential moments, where you&#8217;re able to, say, hear what reverb sounds like, I&#8217;ve been able to design experiences where you can show it dry, meaning without that particular effect. Let the user hear it for themselves, let them understand it. And, ideally, let them play around with it, take a knob or a dial to pour more in or take more off, to really start to feel it.</span></p><p><span>That&#8217;s just one example of how jargon will impact your experience, because you want to get a sound &#8212; you know the sound, you&#8217;ve heard it, you know an artist who plays in that style or sings in that particular way. How do you get there with words? It&#8217;s impossible.</span></p><h3><span>How do you know whether a design choice is working?</span></h3><p><span>As easy as a heat map, or using triggered events. When they came to this part of their journey &#8212; &#8220;Do you know what reverb sounds like? Here, press this&#8221; &#8212; how many people pressed that? That&#8217;ll tell you right there. And then, following that down, what are the other KPIs you have for that feature? Whether it&#8217;s converting, buying something, or taking on a subscription, you start to see that if you get people to understand and feel and really get what they&#8217;re doing, your KPIs will increase across the board. If they know what they&#8217;re doing, they&#8217;re going to enjoy it and do more of it.</span></p><h2><span>Making the case: Inclusivity is growth</span></h2><h3><span>When inclusivity doesn&#8217;t show up as a metric the way conversion or retention numbers do, how do you justify keeping it on the roadmap?</span></h3><p><span>I justify it by not having it on the roadmap &#8212; by not saying, this is going to be an exclusive experience, because that doesn&#8217;t really resonate in a boardroom. I don&#8217;t mean it literally, but when resources are limited and you&#8217;re trying to hit numbers, that&#8217;s something that&#8217;s not really valued.</span></p><p><span>So what is inclusivity? If you take a step back, it&#8217;s really about growth. That&#8217;s it. If I can say, I will increase our user base by X percent, they&#8217;re like, great, go do that. That&#8217;s inclusivity in practice &#8212; reaching people who may not have previously felt these products were for them and helping them achieve what they came here to do. It&#8217;s making sure we&#8217;re speaking their language and applying learning techniques in a way that&#8217;s meaningful to them. Growth follows when you create that value.</span></p><p><span>At several companies I&#8217;ve worked with, that is a metric we track. We call it the moment of joy, that aha moment. Anyone who&#8217;s trying to make music can hear it. They did it, they get that feeling of joy, and you can really capture that and go, ding, </span><em><span>that</span></em><span> moment. From a product perspective, we want to get you there. We back into the moment you first enter our experience to that moment of joy: How do we get you there, how long will it take, and what are those inflection points we can look at from a data perspective that say, if we follow these steps, then you&#8217;re 7X more likely to have that moment. And we&#8217;ve got you for life. We&#8217;ll be that specific and meticulous about every hour in that journey to get you to that moment.</span></p><p><span>How do we know that? Anyone who&#8217;s a player who&#8217;s been playing for 25 years, we&#8217;ve asked this one question: What was the first song you learned to play? They all know it. They all talk about how they learned it. They all talk about that moment when I figured out that last chord that George was playing &#8212; &#8220;Aha, that was it, I was in.&#8221; Or that moment I sang and heard my voice sound like so. Every musician has that moment, so how can we help you get there quickly?</span></p><h2><span>Leading inclusively, from culture to the next generation of creators</span></h2><h3><span>As a leader, what are your golden rules for building a culture of inclusivity?</span></h3><p><span>Being absolutely inclusive. How can I expect product leaders or designers to be inclusive in all their work if I&#8217;m not inclusive with a team? How can I build a team that has a variety of voices, opinions, backgrounds, and facts?</span></p><p><span>Diversity isn&#8217;t just gender-based or background-based &#8212; it&#8217;s also about what kind of people they are. Are they super type-A, the type who always achieves their goals and likes to make a checklist? If I&#8217;m designing a learning product and I&#8217;ve got all those types of people, it&#8217;s kind of easy. I also need to hear the voice of the person who really wants to do it, but they&#8217;re too busy &#8212; how can I support them?</span></p><p><span>What about someone on my team who&#8217;s a busy parent with kids to work around, but who really wants to satisfy their dream of being able to play an instrument? I want to hear from them too. How do you find time? How can you get 20 minutes out of your day? It&#8217;s about bringing people with different backgrounds, perspectives, experiences, and ways of approaching the world &#8212; and creating space for those perspectives to shape the work.</span></p><h3><span>As the definition of a musician keeps stretching, what&#8217;s the next kind of creator most companies in this space haven&#8217;t noticed yet?</span></h3><p><span>I don&#8217;t know about &#8220;haven&#8217;t noticed.&#8221; There&#8217;s a tremendous amount of attention right now on people who may not identify as traditional musicians but are creating music. If you&#8217;d asked me that question a year ago, I would&#8217;ve answered the same way. But as of now, the spotlight is on that cohort, on those potential users, with such a flurry of excitement. And I love it &#8212; I&#8217;ve always believed everybody should be able to make music, even though a lot of people in the industry are resistant to that. This very underserved, huge user base is now getting the love and attention it deserves.</span></p><h3><span>What was your aha moment when you fell in love with music?</span></h3><p><span>I get the chills even thinking about it. I am what I like to call a perpetual beginner &#8212; I&#8217;ve never really learned how to play, but I do remember being a teenager and wanting to be in a band because I loved music so much. My aha moment was that time when I was in a garage with my friends and I became the lead singer because I didn&#8217;t have any instruments, and it was easy. I could do it. But having that moment where you&#8217;re singing and the drummer&#8217;s playing the drum part and the bass is playing, the guitar is going, and suddenly &#8212; we&#8217;re a band! It was so exciting. There&#8217;s nothing like that feeling of playing together and making a noise that sounds at least somewhat similar to the sound you&#8217;re trying to make. It&#8217;s amazing.</span></p><p><span>It&#8217;s so much fun. It&#8217;s frustrating and hard, but that&#8217;s why I feel like I&#8217;m in a good place to support that user &#8212; I&#8217;ve made a career and a life of music without really being able to play. So you can do it. Anybody could do it. It&#8217;s true, especially now.</span></p><h3>What does LogRocket do?</h3><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/?substack">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[Leader Spotlight: Making AI disappear into the flow of work, with Swapna Oundhakar]]></title><description><![CDATA[Swapna Oundhakar is a product leader who builds enterprise AI for the workplace.]]></description><link>https://stories.logrocket.com/p/leader-spotlight-swapna-oundhakar</link><guid isPermaLink="false">https://stories.logrocket.com/p/leader-spotlight-swapna-oundhakar</guid><dc:creator><![CDATA[Katie Schickel]]></dc:creator><pubDate>Thu, 10 Sep 2026 05:02:56 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!WMA0!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><em><span>Swapna Oundhakar is a product leader who builds enterprise AI for the workplace. Most recently, she was Vice President of Product Management at Kore.ai, where she led an agentic AI portfolio spanning the employee lifecycle across HR, IT, and recruiting, shipping autonomous agents on Slack, Teams, and the web for large enterprises. Swapna began her career in technology transformation consulting at Capgemini. She spent more than a decade at ADP, ending as Director of Product Management, where she built the first version of ADP Workforce Now, now a $1B+ revenue flagship. Before her most recent role at Kore.ai, Swapna also led Amazon&#8217;s global career development platform for a distributed frontline workforce and ran the enterprise payroll and tax portfolio at Business Software, Inc.</span></em></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!WMA0!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!WMA0!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!WMA0!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!WMA0!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!WMA0!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!WMA0!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png" width="895" height="597" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:597,&quot;width&quot;:895,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1326736,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://stories.logrocket.com/i/213577856?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!WMA0!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!WMA0!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!WMA0!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!WMA0!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5b3e20aa-f3ed-4ef7-a92f-3aea2da2bca1_895x597.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em><span>In our conversation, Swapna talks about designing AI around the work employees are already doing rather than asking them to adopt new habits. She discusses building trust in nondeterministic systems, the new evaluation practices AI products require, and why the future of enterprise AI lies in agents that can carry work across systems.</span></em></p><div><hr></div><h2><span>Making enterprise software disappear</span></h2><h3><span>You&#8217;ve built products across payroll, HR, recruiting, and workforce technology. What&#8217;s the biggest lesson you&#8217;ve learned about what employees actually need from the software that they use every day?</span></h3><p><span>Employees just want to get the job done. They need the software itself to disappear. Nobody wakes up wanting to use an HR system &#8212; they just want to get paid correctly, get answers to questions they might have about policies, get a laptop replaced, that type of thing. Their problems have essentially remained the same despite changes in technology, and I&#8217;ve been through those technology transformations. People want to use the software to do what they need to do and then be on their way.</span></p><h3><span>With Kore.ai&#8217;s AI for Work platform, what problem were you trying to solve beyond simply adding another AI assistant to the workplace?</span></h3><p><span>This is true for any enterprise: the biggest problem employees have is needing to use 10 different systems to get their jobs done. Ideally, they should be able to go to one place to ask their questions and get the help they need.</span></p><p><span>The first question is, &#8220;Where do I need to go to do that stuff?&#8221; As everybody starts to add AI, the problem compounds. Even if I need to go to HR, do I need to go to a specific HR system for policy questions? And another for benefits? As an employee, that&#8217;s not a good process.</span></p><p><span>The problem that I was trying to solve is how to make employees&#8217; lives easier using AI. We need to make it simple for people to do their work without having to go to 10 different systems. This starts with meeting them where they already work. That could be in Teams or Slack, or a web UI. We don&#8217;t need to make people develop a new habit.</span></p><h2><span>Building trust into enterprise AI</span></h2><h3><span>Enterprise AI operates in regulatory environments where mistakes can have real business consequences. How does that change how you approach the product design, testing, and rollout?</span></h3><p><span>Even in enterprise software, compliance has to come first. I&#8217;ve built those products from the ground up. Mistakes in payroll, compliance, and HR have huge consequences. It&#8217;s the same in AI, and that risk compounds even further.</span></p><p><span>You need to make sure answers are thorough and that humans are in the loop in sensitive areas. You need to be very careful about preventing hallucinations, providing accurate answers, escalating to humans when needed, and accounting for compliance-level rules. The overall compliance landscape should be thought through before you ever push something live. Building specific guardrails &#8212; like masking sensitive data, for example &#8212; is not an afterthought. It needs to be included in the design itself.</span></p><p><span>When you&#8217;re doing regression testing, you should have golden datasets and rules that you&#8217;ve already thought through. What types of policies and procedures will inevitably come up? These are called evals in AI. When you&#8217;re rolling the product out, you have to start with context grounded in the organization&#8217;s data so that the agent doesn&#8217;t hallucinate. If it can&#8217;t answer, there need to be correct fallbacks so that it doesn&#8217;t make things up. It should either escalate to a human or say, &#8220;I don&#8217;t know the answer to this question, but here&#8217;s a document for you to review or here&#8217;s someone you can ask.&#8221;</span></p><h3><span>How do you mitigate the hallucination risk of LLMs and get the end user to trust an AI agent with sensitive data?</span></h3><p><span>You have to build trust from the moment a product is rolled out. As a user, the first time I interact with a system, I expect it to do the job I want it to do &#8212; almost as well as a human does. If it doesn&#8217;t know the answer, it should send me to a human rather than make one up.</span></p><p><span>A company&#8217;s policy document repository is very important here. For example, I&#8217;ve worked with customers who have laid out &#8220;moments that matter.&#8221; In an HR scenario like a bereavement leave case, LLMs are very good at handling that type of sensitive information. But as soon as the situation involves sensitive data or something like an employment violation, those are the &#8220;moments that matter.&#8221; Once an AI detects that type of intent, it should give a very empathetic answer and pass the situation along to a human because mistakes in those contexts are very costly.</span></p><p><span>When you&#8217;re designing AI workflows, you need to understand those workflows end-to-end: what are the connection points between different systems and different users, and how do I now redesign and rethink this entire workflow while keeping humans in the loop at specific moments?</span></p><h3><span>What new practices have become essential for evaluating an AI product before it ships?</span></h3><p><span>One of the biggest things is evaluations &#8212; making sure the product is working correctly. There are two types: one happens during the build itself, and the other happens during internal rollouts or pilots. During the build, you anticipate what people might ask, but when you actually roll the product out and people start asking questions, you may realize that you haven&#8217;t thought of every scenario.</span></p><p><span>What&#8217;s different here is that you&#8217;re not evaluating a UI; you&#8217;re evaluating conversations. There&#8217;s no UI to speak of, so the design practices become different. Are you empathetic? It&#8217;s a whole new way of thinking. There also isn&#8217;t the single point of failure that you used to have in traditional software development. It&#8217;s generative so that the answer can be different every time. Pinpointing where something is going wrong requires investigating a whole range of possibilities: &#8220;Is it in retrieval? Where did it actually go wrong? How do I even fix this?&#8221; There&#8217;s not a straight answer.</span></p><p><span>It&#8217;s very different from traditional UI development because now you are dealing with conversations or voice. Traditional development practices are completely thrown out of the loop. Still, spec-driven and test-driven development concepts, which have existed in the software world, are now reflected in practices like evals and spec-driven development. They&#8217;re new practices that you need to adopt.</span></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive new posts every week.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h2><span>Rethinking product development for agents</span></h2><h3><span>Having worked in both traditional software and this non-deterministic new world, do you have a preference?</span></h3><p><span>Currently, I do not know how I would go back to the old way. Now, I&#8217;m thinking in terms of agents and how to develop them. When I&#8217;m given a problem, I try to map the workflow, and my mind immediately goes to, &#8220;What agent would solve this workflow? What systems are involved?&#8221; My thinking has changed dramatically. I&#8217;m also thinking in terms of conversations. To paraphrase Satya Nadella&#8217;s term, AI has become the new UI. I find it difficult to go back to traditional software development at all.</span></p><h3><span>Your team rolled out the platform internally first. What did that reveal that you couldn&#8217;t have learned through customer interviews or usability testing alone?</span></h3><p><span>Initially, I was rolling it out to different users. We spoke to different user groups, really trying to chart out their days: &#8220;What&#8217;s your day-in-the-life like?&#8221; These are UX research practices that have existed for a long time in software development. Those practices are still important. Nothing substitutes for good UX and UX design, though the traditional terms have changed a bit. Talking to users, talking to customers, and finding the right problem to solve have not changed at all.</span></p><p><span>I spoke to our HR, IT, and sales departments to learn what their jobs-to-be-done are. Once we had a good set of interviews and a grasp of our internal systems and workflows, we evaluated whether the use cases we built were still applicable. What tweaks would we actually need to adopt it within the enterprise? We had the golden datasets, and we&#8217;d thought through the test sets, but when it came to rolling it out for real, there were surprises. People use every system differently, so we didn&#8217;t anticipate certain questions.</span></p><p><span>Like any software rollout, we learned as we went, adapted, and evaluated. It&#8217;s true for every single enterprise: Who are the users? What problems are they trying to solve? What systems are they touching? What workflows are they using? Then, after rollout, does it work? Have people actually adopted it? Are they coming back? Is it successful?</span></p><h3><span>Can you share an example of something that surprised you in the rollout?</span></h3><p><span>Initially, when we were working with policy documents, we had to configure the system so that the right countries&#8217; policies were displayed to the right people. We only realized that during testing. It all came down to how you organize your documents and make sure answers come from the right documents in the right format.</span></p><h2><span>Designing for adoption, not disruption</span></h2><h3><span>AI is changing how people work almost in real time. How do you design a platform that can evolve just as quickly without forcing employees to reinvent their workflows constantly?</span></h3><p><span>You need to meet employees where they work &#8212; whether that&#8217;s in a chat or a web UI, you shouldn&#8217;t change that. The user&#8217;s habit actually lives in those systems. In most cases, it is in Slack or Teams, and anytime you make changes, the experience should survive every model upgrade, every capability change, and every new architecture change behind it.</span></p><p><span>That&#8217;s why the platform-first approach works. Models can be swapped, but you need to be able to do new regression testing. The workflows that people have already gotten used to should not break.</span></p><h3><span>When you&#8217;re building an employee-experience platform, what signals tell you that the product is actually improving how people work?</span></h3><p><span>This has always stayed the same in enterprise software &#8212; it&#8217;s about retention more than adoption. The first metric is whether or not people have taken to it. Then, the real test is whether they are coming back to use it.</span></p><p><span>For example, with HR software, you are often required to use it &#8212; there&#8217;s no other way for you to submit a time-off request. But if it is optional, and now you can submit a time-off request using AI, are people actually using it? Have they come back because they liked it and it made their work simpler? Further, are they now starting to use it for complex tasks? As we move to agentic AI tasks, a good example is a user telling the AI, &#8220;Plan my leave. Check my vacation, look at my policies, and tell me the best leave that I can take that optimizes my time off with my vacation days.&#8221; That&#8217;s a complex-level task that an agent can actually do.</span></p><p><span>In IT, for example, the AI can go in, check when your laptop lease is going to expire, and automatically send emails to an administrator that the laptop needs replacing. That&#8217;s a level of autonomy, but of course, I always stress having a human in the loop.</span></p><p><span>When people use it as their first stop because it&#8217;s that useful, that&#8217;s when you know it is successful. There are some easy metrics to track, such as lower ticket volume, because you can conclude that people are getting their questions answered by the AI. HR can then do higher-level tasks and focus on strategy rather than answering questions.</span></p><h2><span>From conversations to end-to-end work</span></h2><h3><span>Looking ahead, what do you think the best enterprise AI platforms will do three years from now that today&#8217;s products aren&#8217;t yet capable of delivering?</span></h3><p><span>They&#8217;ll be able to carry work across systems instead of just answering questions. You&#8217;ll hear about orchestration in the AI world &#8212; a typical enterprise will have 50-plus systems. There&#8217;s an end-to-end workflow that can execute across those systems rather than relying on these single points of contact with AI tools.</span></p><p><span>Further, today&#8217;s products are mostly conversational. They&#8217;ll summarize documents and answer questions, but the opportunity is to cut across workflows. For example, &#8220;I can manage your end-to-end day from morning to evening. In the morning when you come in, I can tell you what meetings you have. Here&#8217;s your dossier for them.&#8221; Work should be a lot simpler, and it still isn&#8217;t. These systems need to be proactive, anticipating what you need and taking that busy work away. This will make work more enjoyable because AI systems can do a lot of the grunt work.</span></p><h3><span>As expectations for output accelerate alongside AI, how can AI help employees keep up?</span></h3><p><span>It&#8217;s the cause of these increased expectations, but it can also help you. Nowadays, there&#8217;s no other way &#8212; to do your best work, people are being asked to do more work with less. It&#8217;s not stopping. So AI is an essential component.</span></p><p><span>We all talk about using humans in combination with AI. I truly believe in that and think AI should be seen as an asset rather than a replacement. If you view it this way, it can augment people to the point of driving higher satisfaction. That&#8217;s a win-win for both the organization and the employees who work there.</span></p><h3>What does LogRocket do?</h3><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/?substack">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[Leader Spotlight: The context layer is the new differentiation, with Dhwani Soni]]></title><description><![CDATA[Dhwani Soni is a product, design, and transformation leader whose career has spanned enterprise software, security, communications, customer experience, analytics, and AI.]]></description><link>https://stories.logrocket.com/p/leader-spotlight-dhwani-soni</link><guid isPermaLink="false">https://stories.logrocket.com/p/leader-spotlight-dhwani-soni</guid><dc:creator><![CDATA[Jessica Srinivas]]></dc:creator><pubDate>Wed, 09 Sep 2026 07:03:11 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!iDDo!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><em><span>Dhwani Soni is a product, design, and transformation leader whose career has spanned enterprise software, security, communications, customer experience, analytics, and AI. At 8x8, she holds portfolio-level leadership responsibilities across product management, design, research, operations, and AI-powered experiences &#8212; building on earlier design leadership roles at MobileIron, SAP, Dell Services, and Samsung Electronics.</span></em></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!iDDo!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!iDDo!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!iDDo!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!iDDo!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!iDDo!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!iDDo!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png" width="895" height="597" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:597,&quot;width&quot;:895,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1314737,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://stories.logrocket.com/i/214466377?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!iDDo!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png 424w, https://substackcdn.com/image/fetch/$s_!iDDo!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png 848w, https://substackcdn.com/image/fetch/$s_!iDDo!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png 1272w, https://substackcdn.com/image/fetch/$s_!iDDo!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4ed04d49-a04a-4da7-8f6c-5e57548528dd_895x597.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em>Her perspective is grounded in a central conviction: as AI lowers the cost of building software, features alone will become increasingly difficult to defend, and durable differentiation will come from the context, semantic infrastructure, workflow intelligence, and trust systems that let products understand a business and evolve with it. In this conversation, Soni discusses what it means to make an enterprise legible to AI, why products must earn the right to become invisible, how personalization can remain transparent and controllable, and why the products that endure will be the ones that continuously increase their value to customers.</em></p><div><hr></div><h2><span>Beyond features: The new durable moat</span></h2><h3><span>With tools like Claude Code dramatically lowering the cost of implementation, many products are starting to look remarkably similar. What still creates durable differentiation?</span></h3><p>Durable differentiation is moving away from the feature itself toward more infrastructure, context, and the semantic layer that sits underneath it. Features are already copyable. I like to think of them more like menu cards. One vendor introduces a feature capability, the other can copy it fairly quickly, and it&#8217;s another version of it that is sitting with a different prompt, different model, maybe slightly different implementation approach. Now, as the cost of building falls, the shelf life of the standalone product differentiation is reducing significantly.</p><p>What&#8217;s becoming really hard to replicate is the system that enables the AI to understand the business environment, like the relationships between the data, the meaning of the workflow, the users&#8217; context, what decisions are being made. What are the decisions that the customer cares about? What are the outcomes that they&#8217;re driving for? The semantic and the contextual layer is what makes the entire enterprise legible to the AI.</p><p>Once you have that, you can enable the model to make smarter decisions, orchestrate across workflows, and deliver an experience that is relevant to the end user, rather than simply generating a generic response to them or building a generic feature. And the second part of the differentiation is the infrastructure itself. The ability to connect systems, the right systems, preserve the context over a multithreaded set of information, govern the actions, and create the feedback loops, so the product can learn from the usage and the outcomes itself.</p><p>So I believe that the products that will win will be self-evolving systems, honestly. They will not simply ship a fixed set of features. They will continuously become more intelligent and valuable as they understand the customer&#8217;s environment better. Otherwise, we&#8217;ll be just building point solutions with increasingly similar capability, and that&#8217;s really not differentiation. So as the features start to become a commodity, the advantage is the context graph, the orchestration and the infrastructure underneath, and a self-healing, self-evolving system for the product to evolve with the customer&#8217;s needs.</p><h3><span>If feature parity is becoming inevitable, where should teams invest to create experiences that are difficult for competitors to replicate?</span></h3><p>Teams should invest in understanding the customer&#8217;s environment deeply enough that the product becomes connective tissue within it. That starts with listening carefully to the customer &#8212; not just to the feature being requested, but to the outcome they&#8217;re trying to achieve, the systems they already use, the workflows they depend on, and the friction preventing them from realizing value.</p><p>From there, the opportunity is to build the semantic layer and contextual intelligence that let the product fit naturally into that environment, rather than sitting beside the customer&#8217;s infrastructure as another isolated tool. Teams also need to invest in curation and adoption: providing a capability isn&#8217;t enough. You have to configure it for the customer&#8217;s context, help embed it into the way they work, and make sure they&#8217;re realizing the intended outcome.</p><p>A competitor may be able to replicate the visible feature. It&#8217;s much harder to replicate the accumulated understanding of the customer&#8217;s environment, the integrations, the adoption model, and the trust that develops over time. That&#8217;s when a product becomes more than software &#8212; it becomes part of the customer&#8217;s operating infrastructure, and therefore much harder to replace.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive new posts every week.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h2><span>The paradox of invisible design</span></h2><h3><span>Sometimes the best products are those that become &#8220;invisible.&#8221; What does invisibility actually mean from a product design perspective?</span></h3><p>I think it&#8217;s such a powerful moment when a user lets you put your product that is almost invisible to them on their mobile phone. A product earns that right to become invisible when the user trusts it deeply enough to depend on it without having to constantly manage it. My perspective is drawn from my earlier work in security. The best security systems are often the ones that recede to the background, that do the work for you but you don&#8217;t need to worry about them, because they only surface when there&#8217;s a threat or an anomaly or an insight that really requires attention.</p><p>I think the same principle applies more broadly today beyond the security industry. A product becomes invisible when the interface is no longer the center of the experience, but the solution is, the outcome is. The user continues to receive the outcome that they care about, while the system blends naturally into their workflow or the work routine.</p><p>The customer should be able to define the problem, or you sit down and understand the outcome clearly that the customer&#8217;s driving for. The product should almost become so frictionless, the way we deliver it, that it becomes a part of the routine, whether they need an interface, or they need a recommendation, or they need an automated action or no visible interaction at all. What do they care about, and how can we deliver that in a very orchestrated manner, in a more trustworthy manner? That&#8217;s where a product can really be invisible, but I do not mean invisible to mean opaque. The user must still understand what the system is doing, retain control over the important and consequential decisions, trust the product, and understand the business context in which it is operating.</p><p>Ultimately, invisibility is really not the absence of the product, but it&#8217;s an evolution of the experience that you bring to the user and the presence of value, without the burden of operating the product or the product requiring the user&#8217;s constant attention. In short, a product becomes invisible when trust is high, friction is low, and the outcome stays visible even when the interface doesn&#8217;t.</p><h2><span>Personalization without complexity</span></h2><h3><span>As products become more personalized to each user&#8217;s context, how do you keep that from becoming overly complicated or error-prone?</span></h3><p>Personalization, again, has to start with a deep understanding of the problem the user is trying to solve. Personalization should not require users to reorganize their work around the product. The product should adapt to their context and working model. This was valuable in the pre-AI era as well.</p><p>The first principle for a product to get personalized or to be adopted in a personalized manner is to show the work, especially early in the relationship. The AI should make reasoning, assumptions and proposed actions, visible enough such that the user can understand how it reached the conclusion. Over time, as the system improves itself, the user may choose to delegate more authority, but that trust at the beginning has to be earned. That&#8217;s where human-in-the-loop comes in.</p><p>The second principle in personalization is more about inspectability. Let&#8217;s say the product goes more invisible, but the user should always be able to go back, query what happened, why a certain decision was taken, and understand if or why the AI may have taken a wrong turn. It should be reversible, because that&#8217;s what is sometimes needed. That makes the system more stable, because when it&#8217;s imperfect, it&#8217;s not inscrutable.</p><p>The third principle is flexibility, like how I was talking about being able to revert back, but it&#8217;s the flexibility in how the outcome is delivered. Personalization should not be limited to the way that the interface has to change. Different users may need the same outcome, same capability, but delivered to different experiences. It could be a mobile experience, it could be a web interface, it could be an automated workflow, it could be an MCP-enabled interaction. It could be as simple as, &#8220;Pull it into a Google Sheet for me.&#8221; That&#8217;s the flexibility of personalization. That model hasn&#8217;t changed. So it&#8217;s not about personalizing the interface, it&#8217;s about personalizing the experience by leveraging the workflow that works for a user&#8217;s mental model, and it&#8217;s deeply rooted in the user&#8217;s problem. Again, transparent enough to be able to earn the trust, flexible enough to fit the workflow, and controllable enough that the user never feels that they are not in control or the product is unstable. Ultimately, the AI has to earn the right to make decisions on the user&#8217;s behalf.</p><h2><span>Products people don&#8217;t want to leave</span></h2><h3><span>Switching costs are often framed around data lock-in or integrations, but there are also emotional and behavioral switching costs. How do you design products a user doesn&#8217;t want to leave?</span></h3><p>This is actually my favorite question, because the premise hasn&#8217;t changed between pre-AI and post-AI. That has always been the core challenge: how do you build something that people just love to stay in, day in and day out? I think, again, it goes back to understanding their mental model, how they work, how they make decisions, what they trust, and where friction appears for them. The product has to fit naturally in their mental model, rather than asking them to get out of it and continuously adapt again and again to the product that you have built for them, and that&#8217;s challenging. That&#8217;s complete friction right there.</p><p>The second part is trust. Customers stay when the product is reliable, it&#8217;s understandable, and it&#8217;s consistently helping them achieve the outcomes that they care about. Then the third, which is increasingly important &#8212; this is the part that gets me really excited now that it&#8217;s possible &#8212; is evolution. The context today is already evolving. The data is always fresh. It&#8217;s real-time. The priorities are always getting updated. Workflows are getting really flexible and real-time adaptable as well. The insights available to them are getting generated. It&#8217;s learning and it&#8217;s feeding you information. What if the product evolved with the context? The product should not remain static.</p><p>I think the strongest products now will need to become increasingly self-healing and self-evolving. So they will detect where something has changed, adapt the experience or the workflow, automate what can be automated, request permission when appropriate, continuously improve, without requiring the user to go elsewhere to fill minor feature gaps. Because you&#8217;ve got the right context which is constantly updating. You&#8217;ve built it over time. You&#8217;ve built a connective tissue where it works with your infrastructure. All you need is the product to evolve, and you&#8217;ve already set the experience in the way that you want to deliver it. That creates a much stronger form of retention for the customer and customers want to stay. This way the product understands their environment, reduces their effort, and becomes more valuable over time.</p><p>I think the goal is not to create a dependency through lock-in. The goal is to create confidence that it&#8217;s continuously relevant and it&#8217;s working for you, not the other way around.</p><h2><span>Enterprise UX in an AI-native world</span></h2><h3><span>With AI, has the definition of &#8220;consumer-grade experience for enterprise&#8221; changed?</span></h3><p><span>I think if you root the definition of a good enterprise experience in deeply understanding the problem, then you&#8217;re fine either way, with AI and pre-AI. What has really changed is the evolution of what experience means. Good enterprise UX was once largely defined by making complex software easier to use: clearer navigation, more coherent information architecture, fewer steps, and less training. Those qualities still matter, but they are now table stakes.</span></p><p><span>What has changed is the unit of experience. Enterprise UX is no longer confined to a screen or device. It now spans workflows, roles, data, integrations, automated decisions, and interactions across multiple systems.</span></p><p><span>A beautifully designed interface inside a fragmented workflow is not a good experience. If users must reconstruct context, repeat information, move between disconnected systems, or manually coordinate decisions, the underlying problem remains unresolved.</span></p><p><span>AI expands the remit of design further. We are now designing how systems interpret intent, communicate uncertainty, request approval, take action, and learn from outcomes.</span></p><p><span>Good enterprise UX is therefore no longer simply about making complex software usable. It is about designing coherent, intelligent, and trustworthy business systems. Screens still matter, but they&#8217;re no longer the primary measure of the experience.</span></p><h3><span>If voice and natural language become the primary frontend interface, how does that change the role of product design?</span></h3><p>The experience and the way you deliver it is definitely changing. Previously, you had to go gather the intel. You would go to solutions engineers, you&#8217;d go to customers, you&#8217;d set up labs, you&#8217;d look at competitors&#8217; information, you&#8217;d look at documentation. You were gathering context. Now that context is automated for you, so how do you automate the ingestion of context? I like to call it the ExperienceOS. So you&#8217;re now building an ExperienceOS first, the ingestion of that customer feedback, the competitive intel, how products are being used, and behavioral data together, creating an AI-ready to read an automated .md file with full context of the product experience.</p><p>Once you have that, leverage a design system that is AI-ready that can read natural language and an .md file for AI-dev creation, and now you&#8217;ve got end-to-end. You&#8217;ve got context, a prompt, and you&#8217;ve got your natural language input that pulls it all together. Your design is now fed by voice, natural language, and with the context inputs. Again, the work needs to be done on the context layer, semantic layer, pulling information from so many different sources and making sure there is no hallucination. That&#8217;s one place where you put a human in the loop or a learning system there.</p><p>The second is if it&#8217;s spitting out automated experience, whether it&#8217;s screens, whether it&#8217;s MCP, whatever that might be, how is that being checked? There needs to be a person who&#8217;s reviewing that, or an agent that has learned what it needs to deliver. So I think that&#8217;s how product design is evolving, from the overall ingestion of an evolving contextual layer of multiple sources, rooting it into the business outcome, rooting it into the experience that you want to deliver holistically as a company, as a vendor, and then using natural language prompts. A lot of work needs to go on the prompting itself. You do that work, and then it leads to a much more streamlined model of delivering product experience.</p><p>This can make it feel a little like design is not important, but I think a lot of work now goes on building the right systems design, building the right libraries, rooting it into the right decisions such that a single command, even from a designer, can fix the system completely. So a lot of focus goes on the platform layer, context layer, and it&#8217;s not about tweaking one screen at a time. It&#8217;s tweaking the system, making it more intelligent at the same time. The visible frontend may increasingly become language, but the real design material is the behavior of the system.</p><h2><span>What survives the commodity trap</span></h2><h3><span>If every company can generate code, ship features quickly, and access increasingly similar AI capabilities, what will separate the products that endure from the ones that become commodities?</span></h3><p>I think products that will endure are those that are rooted in business outcomes &#8212; products that customers can trust, that they can almost have a relationship with. That&#8217;s because the product is compounding in value for them; self-healing is one example of that. Then why would I buy from you versus somebody else? Because you have an infrastructure that it&#8217;s rooted in, that you understand my business, and that together both of them become a connective tissue. I think that&#8217;s the one that endures.</p><p>Products that become commodities will just provide feature-level capabilities, point solutions. Enduring products will help lead organizations through their own transformations &#8212; continuously providing that transformation to customers is what endures. Commodity products provide capability; enduring products continually expand what the customer is capable of achieving.</p><h3><span>Many valuable product metrics today weren&#8217;t measurable just a few years ago. How is AI changing what product leaders should be measuring?</span></h3><p><span>AI is changing product measurement in two ways. First, it allows us to measure signals that previously existed only in unstructured conversations. Second, it allows us to connect those signals to product behavior and business outcomes. Product leaders no longer have to rely exclusively on surveys, interviews, support tickets, and manually synthesized field feedback. AI can analyze customer conversations, implementation notes, sales commitments, product usage, support interactions, and behavioral data together. That creates entirely new measures. We can identify which promises were made during the sales cycle, whether the product delivered on them, where adoption stalled, which customer problems are becoming systemic, and what interventions are most likely to improve the outcome.</span></p><p><span>These capabilities also make measurement more dynamic. Once an organization identifies and resolves one source of friction, the next constraint emerges. The KPI itself may evolve as the business becomes capable of seeing and solving new problems. That is particularly important for CIOs and transformation leaders. Many know they need to lead change but are looking for partners who can help determine what to measure, interpret what the signals mean, and identify the next action.</span></p><p><span>The most valuable product organizations will not simply report what happened. They will continuously identify what is changing, explain why it matters, and recommend what the business should do next.</span></p><h2><span>As AI becomes more embedded in how products work, how do you see the role of product and design leadership evolving?</span></h2><p>The role of product and design leadership is expanding. For years, we largely designed software that waited for users to operate it. We are now designing systems that interpret, recommend, act, learn, and increasingly participate in the work itself. </p><p>That requires a broader form of leadership. Product strategy, design, AI, customer experience, go-to-market execution, infrastructure, and organizational transformation can no longer be treated as separate disciplines. </p><p>The opportunity is not simply to ship more capability. It is to design the complete system through which a business becomes more intelligent, more adaptive, and more capable. The leaders who shape the next generation of products will be those who can make complexity coherent, make intelligence trustworthy, and make transformation adoptable.</p><h3>What does LogRocket do?</h3><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/?substack">LogRocket.com</a>.</p>]]></content:encoded></item><item><title><![CDATA[Product Sense: What It Is and Why AI Makes It Vital | Kevin Sung, VP of Product (Life360)]]></title><description><![CDATA[Life360's Kevin Sung on why shipping faster raises the price of bad judgment, and how a year of fixing quality problems at Dropbox turned into $24M in ARR.]]></description><link>https://stories.logrocket.com/p/product-sense-what-it-is-why-ai-makes-it-vital-kevin-sung</link><guid isPermaLink="false">https://stories.logrocket.com/p/product-sense-what-it-is-why-ai-makes-it-vital-kevin-sung</guid><dc:creator><![CDATA[Imane Rharbi]]></dc:creator><pubDate>Tue, 08 Sep 2026 13:48:00 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/cbc7d2f3-b5ae-4e0e-885a-21a6ed34084e_895x597.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div id="youtube2-EGApwwmnJQ8" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;EGApwwmnJQ8&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/EGApwwmnJQ8?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><div class="pullquote"><p><em><strong>Listen on:<br><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8">YouTube</a> | <a href="https://open.spotify.com/episode/3hYLYPCTo0DrmTL6ujVOTa">Spotify</a> | <a href="https://podcasts.apple.com/us/podcast/product-sense-what-it-is-and-why-ai-makes-it-vital/id1733103005?i=1000788461756">Apple</a></strong></em></p></div><p><span>Everyone says good PMs have product sense, but how many people can truly define what good product sense is?</span></p><p><span>In this episode, we&#8217;re joined by </span><a href="https://www.linkedin.com/in/kevinsung/"><span>Kevin Sung</span></a><span>, VP of Product at Life360, where he runs ecosystem strategy, the foundational infra and developer experience group, and works alongside the AI platforms team. Before Life360, Kevin spent several years at Dropbox, leading the Usability team that owned performance, reliability, and CX.</span></p><p><span>Kevin&#8217;s worked in both the goal-seeking world and the customer-obsessed one, and he&#8217;s clear about which one AI is about to make obsolete.</span></p><p><span>In this episode, Kevin shares:</span></p><ul><li><p><span>Why product sense is customer centricity</span></p></li><li><p><span>The difference between lazy product management and rigorous product craft (hint: it&#8217;s how strong your opinion is before you start)</span></p></li><li><p><span>How his team at Dropbox found $24M in ARR hiding in a cohort nobody was looking at</span></p></li><li><p><span>And how to get leadership to fund the unglamorous quality work that never wins a roadmap fight</span></p></li></ul><div><hr></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://stories.logrocket.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Product: Behind the Craft! Subscribe for free to receive weekly posts and podcast episodes.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><h2>1. If your job is running every permutation, AI already does it better</h2><p>Kevin isn&#8217;t in the doomer camp on AI and PM roles. But his reasoning for why the job survives is sharper than the usual reassurance. </p><p>AI removed engineering headcount as the bottleneck on learning &#8212; you can prototype and validate in days instead of quarters. That doesn&#8217;t reduce the need for judgment.</p><blockquote><p>&#8220;In a world where you can now ship lots and lots of ideas quickly, it becomes even more important now to have really strong product sense and customer centricity in order to decide that you are shipping the right things.&#8221;</p></blockquote><p>He&#8217;s skeptical of the 10x-your-experiments crowd for a specific reason: someone is on the other end of all those experiments, and a product that changes on you every day makes them a test case. Then comes the line that should make a few PMs uncomfortable:</p><blockquote><p>&#8220;If all you&#8217;re doing is facilitating maximalist tests, an AI will always outperform you, because they will be able to come up with every permutation and actually ship it. So then you are actually obsoleting your own job.&#8221;</p></blockquote><p><strong>Product takeaway:</strong> Audit what part of your team&#8217;s week goes to generating and coordinating options versus deciding between them. The first half is getting automated. If your PMs can&#8217;t articulate a customer-grounded hypothesis before an experiment runs, you don&#8217;t have a product function &#8212; you have a permutation engine.</p><div><hr></div><h2>2. Lazy product management vs. Product craft</h2><p><span>Kevin has a refreshingly straightforward definition of product sense:</span></p><blockquote><p><span>&#8220;What is product sense? It is customer centricity. It is understanding a customer&#8217;s painful needs and being able to translate that into business value [...] One of the big differences between the two is the strength of your opinion, and that opinion is based on your understanding of what a customer&#8217;s needs are.&#8221;</span></p></blockquote><p><span>The rigorous version sounds like this: based on what I understand about this customer&#8217;s life, here is a painful problem, here is my hypothesis for solving it, here&#8217;s how their behavior should change, here&#8217;s the business impact. Then you find out whether you were right &#8212; and either way you get smarter.</span></p><p><span>The lazy version outsources all of that to volume. Kevin points out this predates AI entirely; it&#8217;s the same instinct that had teams running a hundred A/B tests a day and Google testing shades of blue. AI just made it cheaper.</span></p><p><span>There&#8217;s a second-order risk too. If everyone lets the same handful of models generate their ideas, everyone&#8217;s product converges.</span></p><p><strong><span>Product takeaway:</span></strong><span> Make the hypothesis a required artifact, not an optional one. Before anything ships, someone should have written down what they believe about the customer and what behavior change would prove them right. That document is the difference between learning something and generating noise &#8212; and it&#8217;s the part a model can&#8217;t write for you.</span></p><div><hr></div><h2>3. Don&#8217;t let data quietly become a spreadsheet maximization game</h2><p><span>At Smule, Kevin led growth in an environment he describes candidly as goal-driven: here&#8217;s a number, find twenty ways to move it. When a key metric dropped, they&#8217;d spin up a tiger team and build a list of hypotheses.</span></p><p><span>And then:</span></p><blockquote><p><span>&#8220;Seven times out of 10 the culprit would be found by ultimately one of us running through the app itself, going through a flow, and being like, &#8216;Oh, this is clearly the thing that was broken.&#8217;&#8221;</span></p></blockquote><p><span>His conclusion isn&#8217;t anti-data. It&#8217;s about what happens when data is the </span><strong><span>only</span></strong><span> input.</span></p><blockquote><p><span>&#8220;If all you do is look at data and you use data to then double-check other data and form hypotheses on that, and you&#8217;re not talking to the customer, you&#8217;re not experiencing the product yourself, you can very easily just turn this into a spreadsheet maximization game.&#8221;</span></p></blockquote><p><span>Dropbox showed him the other model &#8212; real-world Wednesdays, a large user research team, an expectation that you talked to two to six customers a week. He also makes a sharp distinction about growth work itself: growth is gasoline on a fire. If the product genuinely solves a painful problem, growth accelerates it. If it doesn&#8217;t, you get a pop and a drop.</span></p><p><strong><span>Product takeaway:</span></strong><span> Put a standing constraint on your team that every metric investigation includes someone actually using the product through the affected flow. It&#8217;s the cheapest debugging step available, and it&#8217;s the one most often skipped &#8212; and if your instrumentation can&#8217;t tell you where users are struggling without a manual walkthrough, that&#8217;s the real finding.</span></p><div><hr></div><h2>4. Resentment compounds over time</h2><p>At Dropbox, Kevin led a group called Usability, spanning key user journeys, service performance and reliability, and the entire out-of-product experience &#8212; because your relationship with a product includes every moment you&#8217;re outside of it trying to get help. The team spent a year making lots of small things better and produced roughly <strong>$24 million in ARR from churn reduction alone</strong>.</p><p>The hypothesis they pitched it on was wrong. </p><p>They&#8217;d argued that first impressions matter most, so the gains would show up in month-one and month-two retention. Those numbers didn&#8217;t move at all. What moved was month 13+ &#8212; fifteen years of accumulated customers, improving by roughly half a percentage point. At that population size, half a point was eight figures.</p><p>Kevin&#8217;s explanation for why is the part worth stealing. Their time-to-visual-complete metric averaged around 15 seconds globally, and every one of those slow loads was a small deposit:</p><blockquote><p>&#8220;These things create moments of what I call moments of resentment. Like, man, all my files are stuck in here. I can&#8217;t remove them, but I&#8217;m frustrated, but I can&#8217;t leave you. And then it just builds up and builds up and builds up.&#8221;</p></blockquote><p>Frustrated-but-locked-in users don&#8217;t churn on the day they get frustrated. They churn on the inciting incident &#8212; an outage, a trust-breaking event, a price increase. The price increase doesn&#8217;t cause the churn. It collects on years of resentment.</p><p><strong>Product takeaway:</strong> Your loyal users absorb the most accumulated friction, and their churn stays invisible until something triggers it all at once. Before your next price change or migration, ask what the resentment balance looks like on your longest-tenured cohorts &#8212; a small lift on a very large denominator usually beats a large lift on a small one.</p><div><hr></div><h2>Chapters</h2><p><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8"><span>00:00</span></a><span> Introduction<br></span><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8&amp;t=246s"><span>04:06</span></a><span> Defining "product sense"<br></span><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8&amp;t=465s"><span>07:45</span></a><span> How the PM role is evolving<br></span><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8&amp;t=620s"><span>10:20</span></a><span> Kevin's growth lessons from Smule<br></span><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8&amp;t=862s"><span>14:22</span></a><span> The $24M retention bet<br></span><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8&amp;t=1109s"><span>18:29</span></a><span> Resentment builds until users snap<br></span><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8&amp;t=1358s"><span>22:38</span></a><span> Selling a bet you can't model<br></span><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8&amp;t=1476s"><span>24:36</span></a><span> Enshittification and paper cuts<br></span><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8&amp;t=1671s"><span>27:51</span></a><span> Life360, where trust is the product<br></span><a href="https://www.youtube.com/watch?v=EGApwwmnJQ8&amp;t=1943s"><span>32:23</span></a><span> Conclusion</span></p><h2>Links</h2><ul><li><p><a href="https://www.linkedin.com/in/kevinsung/">Kevin&#8217;s LinkedIn</a></p></li><li><p><a href="https://www.life360.com/">Life360</a></p></li></ul><div><hr></div><h2>What does LogRocket do?</h2><p>LogRocket&#8217;s Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at <a href="https://logrocket.com/">LogRocket.com</a>.</p>]]></content:encoded></item></channel></rss>