Tom Censani is currently Head of Product Design & Research at Optimizely. Previously, he joined ticketing platform Eventbrite as the first UI Designer, where he spent 13 years in various leadership roles growing the design team.
In our conversation, Tom talks about how his web development background has helped him have a better dialogue with technical employees and the people building the product. He shares the structure changes he’s witnessed to the design function over the years, as well as how AI is enabling everyone to be a “maker,” not just engineers and product people.
Designing with a web development background
Your background is primarily in computer science, and you are now involved primarily in product design and research. What was that transition like?
It’s quite a shift. I have a Bachelor’s degree in Computer Science, but about midway through that program, I took a big interest in web design. That was the beginning of the transition from coding to designing for web development. From there, I got into UI interaction, UX, and, later, more ubiquitous product design.
Having that background has granted me a very logical approach to design solutions — not just for the customer, but also for the people who push these decisions live to production. I’ve never been an abstract designer — with every single design that I put forth, I’ve approached it from a lens of how to have that conversation with engineering, and that’s been a superpower.
For example, at Eventbrite I was working on a check-in application for our customers. The previous execution required them to tap each table cell twice. I thought, “That’s very cumbersome. You miss the target a lot.” I asked the engineers if we could do a swipe interface instead, which was a new UI concept at the time. They said, “That’s super hard and new — I don’t think we’ll be able to do that.” With my background, I poked around prototyping sites and found an actual demo. I shared it with the engineers, and by the next day, we had a working prototype in our product experience.
My background allows me to truly talk the talk, and that allows me to have a better dialogue with the people who build and execute the product.
When people think about product design, they often start by thinking about aesthetics. Because you have the perspective of the actual software programming, do you think more about the feasibility and function as well?
People always talk about design as how the product feels, but feeling isn’t just visual — it’s also performance. And that’s where having that language and vocabulary with your cross-functional peers helps you excel.
I think with AI, the days of static work may be numbered. It’s all going to be very collaborative, interactive, and prototype-centric. That lends itself to people who are comfortable with iteration and not necessarily with perfection. Even in the creation aspect, product, engineering, and design can all jump to AI prototyping tools and generate something, and once they’re happy with it, can push it to production. That’s the future — where everyone’s a maker and, therefore, a designer or an engineer. That gray area is terrifying and exciting at the same time.
Experiencing variability and structural change
Over your career, you’ve worked at both small startups and large companies going through an IPO. Are there aspects of the design and research space that worked in a startup but not at a broader scale, or vice versa?
At Eventbrite, we were working on attendee profiles. I invited two of my friends who had just helped launch Twitter’s profiles, and that was one of the biggest step changes that the company experienced. I brought them in to talk about their process and research, and we realized that we were facing the exact problems in terms of what the customer wanted and what the business was hoping to achieve. Yet, the order of magnitude was wildly different.
If you start talking to different businesses or teams, you often find that you’re trying to accomplish the same goals. The variables are often centered around how the company is organized or how you present yourselves to solve those things. That was a fun moment for me — I went from the first UI designer at the company to helping IPO. With all of those changes, there was a lot of variability in how the team was organized relative to how we approached those situations.
You mentioned variability in team organization. Can you expand on how team structures have changed over time, and how you see organizations approaching those changes now?
When I joined Eventbrite, product design was an emerging term. We still had UX architects and visual interaction designers, and in time, those roles dissolved away. Sometimes, the designer is more illustrative and has a wide breadth of work, and other times, they’re hyper technical in a T-shaped role. It was always ever-changing, and it will continue to be.
Now, you have physical designers who work on perfection, where digital ones are never finished and always iterative. You also have data designers and AI designers nowadays, where the interface is eroding, and they’re working more on the experience or the feelings for the user, getting them to a point where they can fully trust the platform. It’s getting more metaphysical and abstract in terms of the responsibilities of design, let alone research, PM and engineering.
Moving forward and the importance of transparency
Part of your tenure at Eventbrite was during COVID. Your work was around ticketing for in-person events, and suddenly, that all went away. What was it like having to redefine the product in a crisis like that?
It was a really challenging time. Suddenly, we had governments telling everyone they can’t gather anymore, and when our entire ephemeral inventory was based on gathering, that was quite a shock.
We got leadership together to prioritize. We had two sides of this marketplace — consumers who aren’t going to the event and need their money back or want to defer it for a later date, and businesses who need to stay afloat. A number of initiatives came out on the finance and transaction side in terms of handling those chargebacks, helping creators, providing them the documentation they need to keep running, etc. We had to be really clear and transparent in the product. We put up virtual banners and sent email campaigns to help them get set up.
Then, there was the moving forward aspect of it, which is shifting to virtual events. There was a time when we explored having our own virtual solution, but we decided to double down on our existing integration and partnership with Zoom. We were suddenly helping people transition from hosting their events live to hosting them virtually. It was a weird time, with stand-up comedians doing virtual improv shows, for example, but everyone was still moving forward. They were working with their community to host events and do what they needed to do to keep going.
You touched on the element of transparency. How did that continue to play a role in product development?
It was important for us to remain as transparent as we could along the way. We accepted that we won’t know a lot of things. And the most important thing to do in that situation is to give some sense of timeline to it. Sometimes, that can feel risky because you’re bringing other people into the uncertainty. But I feel that when you’re able to communicate with people, you’re at least bringing them along in a way that they can appreciate and respect.
COVID was the impetus for a lot of distributed work, and many companies have since maintained their remote policies. How are you approaching building trusting relationships on those teams when they’re distributed virtually?
Thankfully, Eventbrite and Optimizely are both globally distributed companies. At Eventbrite, we had a few offices around the world, and I think it was important to recognize not a hub and spoke model, but more like a cluster model. There were specific strengths in each of those regions, and they had their own cultural sense of self, while also adhering to the company’s global standards.
Having a hiring manager or site representative is key so that there are people who can advocate for that group. They know who to hire to continue building that sense of a team connection. Having that person who can represent the group also gives them the autonomy to facilitate that.
When I came to Optimizely, we were a remotely distributed organization with a few offices. I’ve found success in this model as long as there’s a clear sense of the strategy. Hire the right people to have autonomy and a growth mindset.
How AI closes the gap across organizations
You touched on AI toward the beginning of the conversation. From the perspective of someone who’s done a lot of designing and prototyping from scratch, how do you see AI’s capabilities compared to what you could do yourself?
The TL;DR is that production-ready, scalable, and enterprise-ready modern experiences are all coded, and AI is simply not there yet to match that sophistication. If it gets to that eventually, that will be the big step change that we need.
With that said, what works well is not starting from a blank canvas. Teams need something to discuss and start from. Even if it’s not production-ready, this is where AI is very helpful — it provides a starting point. If there’s an area that a product person wants to explore, they can go into an AI tool like Loveable or Bolt with a prompt and generate an experience to iterate from.
I’m excited by the potential to create the ground floor for a lot of other people at organizations to do this, too. I don’t think design should be a sacred thing, and I often chuckle at the phrase “design thinking.” Everyone thinks and designs, so it’s not a unique differentiator for product design or digital designers. Rather, it’s exciting to see people throughout an organization working through ambiguity, seeking opportunities, and bringing them back to the organization.
Do you have a framework for ensuring that the ways you’re implementing AI are still moving you toward a larger strategic goal?
Everyone is trying to be on the edge of AI right now, but as long as you’re solving for the needs of the customer, you’re moving in the right direction. At Optimizely, we’re thinking about how AI can give customers an experience they didn’t know they needed. What are the net new functionalities that make people really want to use our product? Also, what are the gaps in the product now that can be solved with AI in a better way?
I believe that as long as a product is grounded in those user needs, customers will always see a value and a purpose to it. When you’re seeking solutions to problems that don’t exist, that’s when you run into issues.
As a leader and mentor of people who are working with you, do you have an approach for how you talk to people about embracing a sort of constant impermanence?
Everyone is behind, and that’s a funny thing. It’s exciting to be really thinking about marketing tools and AI right now because everything feels so frontier, which is exciting and terrifying.
When it comes to embracing ambiguity, it’s all about encouraging and creating a space for tinkering. Some organizations can’t do much with AI for legal reasons, and that’s an interesting limiting factor. But I’m curious about how that’s going to break down and change for those sectors because they’re at a point where they need it.
At Optimizely, I encourage people to look at tools and set up time for experimentation. See what it does and what it can change. Who knows, we might emerge with a brand new process and workflow for our operations that propels us forward.
It’s also important to be clear about what these technologies shouldn’t do. I frequently talk to my product counterparts to put up those boundaries and parameters. Having those discussions, as well as encouraging teams to go ahead with exploration and experimentation, will potentially reduce the number of speed bumps.
What does LogRocket do?
LogRocket’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.