From pixels to profit: The next era of UX tools

Moderated by Rich Wong, Partner at Accel, the Insight Out 2025 panel brought together Jonathan Widawski, Founder and CEO at Maze, Ryan Glasgow, Founder and CEO at Sprig, and Alfonso de la Nuez, Entrepreneur, Growth Advisor, Investor, and Author.

To start, could you give us an overview of what your company does and what makes it distinctive?

Jonathan Widawski: My job used to be terrible. I was leading design research, and my job consisted of going to clients and saying, "You have two options. You can build the wrong thing right now—it's going to take two weeks—or we can figure out what to build in the next three months by running extensive interviews."

That tension between the speed at which companies build and the speed at which companies learn is not unique to agencies. It's something that every product organization has experienced. That is the problem that we're trying to solve at Maze. We believe that companies shouldn't have to choose between building fast and building right, and we built a platform that allows organizations to have access to user insights at the speed of product development.

Ryan Glasgow: I'm the founder and CEO of Sprig, and my background is in product management. What I noticed as a product manager at Weebly and other tech companies was the pace of product development going faster and faster every year. How can research keep pace? At the time, the research team was using Qualtrics and sending out 20- to 50-question surveys on a quarterly basis.

So we started Sprig to really take survey and quantitative research and integrate it into the product development life cycle.

Alfonso de la Nuez: I'm the ex-co-founder of UserZoom, which started in 2007 as a platform to help scale research. Before UserZoom, I co-founded a usability testing consultancy, and we spent seven years in a physical lab. That led us to what we wanted to do, which was be more efficient and scale user research.

UserZoom grew to be this remote, online "lab in the cloud"—an automated, scalable way to do research. Then in 2022, UserZoom was acquired by a private equity investor, Thoma Bravo. One of their theses was, hey, if we also buy UserTesting, we can put them together and it becomes a really powerful player. And that's what they did. I left the company for about a year, but then I decided to come back and help out the team, which is what I'm doing right now as an advisor.

This brings us to the AI trend. Do you think AI is underhyped or overhyped?

Jonathan Widawski: First, I think it's still very underhyped, especially for the world of research. When people think about the impact of AI on research, they think about the tactical implementation of it—AI can summarize stuff and create themes. All of this is interesting, but the reality is it's a much bigger shift.

In a world where everyone can build anything, the only thing that matters is what to build. All of a sudden, if companies have been shaped to be time-to-market, output machines for the past 20 years, those things become irrelevant. It commoditizes the capacity for people to build stuff. In that world, insights become the only remaining differentiator. How deeply and how continuously you understand your users becomes the only thing that matters. That's why it's so exciting to be in the research space right now; the value of research is expanding dramatically.

Alfonso de la Nuez: It depends on the day. I've lived through multiple so-called revolutions and crises, so I tend to be a little careful when it comes to new technologies that are changing the world.

However, in this case, the more I spend time trying to understand AI, the more I understand that it's really fundamental because of the capabilities. I don't think anybody knows how crazy things are going to get, but what I can say is that when you apply it to our market, there's certainly potential for optimization and disruption. For me, the obvious one is the analysis. When you have so much data, the biggest pain for most users is spending time analyzing and making sense of it. So I welcome that technology.

AI is going to have impacts up and down the value chain. Where do you think the greatest impact will be for the ecosystem?

Ryan Glasgow: We made a bet very early on that we could solve two key problems. The first was collecting in-context quantitative research data with targeted SDKs. The second was, can we summarize and do open-text coding in the same way as an expert user researcher? This was in 2019. We made a very early bet into the AI field, and we've always seen synthesis and analysis as the starting point for how we can incorporate AI and accelerate our work.

We now have companies collecting over a million responses in a single survey, and AI is able to do the same work at a very close level to an expert user researcher—about 90-95% accuracy. As we learn how to incorporate AI, it's going to move up the process and accelerate all aspects of research, not just synthesis and analysis, but also delivering the research and initiating the research process as well.

We're seeing a shift from massive, six-month-long research projects to more continuous research. What's the portfolio or mixture of how you see customers using your products?

Jonathan Widawski: It's a mix of both. It really depends on who the user of the platform is. We have a platform where the common denominator is the person with the lowest maturity in the organization, so they can autonomously capture and make sense of insights. But we also have professional researchers using the platform who run extensive projects.

I think a trend we're seeing across all personas—designers, product managers—is that speed remains one of the most significant things. Research lives by the product cycle. As product cycles accelerate again and again, you're going to see more focus on research being able to go at least at the speed at which product is being developed. 

Given all this change, what's your career advice for people in this space? How do you prepare for the next 10 to 15 years?

Alfonso de la Nuez: I see two things you can do. One is, clearly, learn about AI. I think it was the Nvidia CEO who said if you're not necessarily an expert in AI, at least learn how to use it. It reminds me of the World Wide Web back in the day when you had to learn how to use a browser.

Beyond AI, I've always advised product people—designers, researchers, and developers—to learn more about business. Business KPIs, what the board cares about. I used to ask developers in our Barcelona office, "Do you know what the gross retention rate is for UserZoom?" As a SaaS business, you know how huge that is. Many had no clue. My take is this industry would benefit tremendously from being able to speak the business language. You're closer to the C-level, you're closer to getting a budget, and you're closer to being heard.

Jonathan Widawski: Soft skills are more important than ever. As the tactical parts of research become mostly automated, the thing that will differentiate great researchers from the rest is the capacity to educate the organization and build influence. There's nothing worse than research that dies in a PDF. The goal for the researchers of tomorrow is to create cross-collaboration, help people drive research, and build influence with executives by understanding how to talk the business language and how to surface insights in a way that people can and want to consume.

Ryan Glasgow: We often hear designers and researchers ask how AI will replace the work they do. What we've seen is that teams are able to stay ahead by thinking about the business impact. If you see AI as a way to accelerate your impact, it's about aligning your work with the business. Research teams are making sure they're not just answering, "Hey, what color should the button be?" but instead looking at the quarterly objectives or what the CEO is focused on. How can you move the needle for critical business metrics like onboarding conversion or customer churn? If you think about AI as a technology that allows you to increase your impact, it will allow your team to be more transformational for the business.

A common argument is that research slows down product development. Is that a fair characterization?

Alfonso de la Nuez: Am I the only one that thinks that research actually makes it faster than slower? Anyone? It's not that many hands, guys. Come on. Why? Because you avoid mistakes. There's no way a designer and developer is going to get it right. If you can stop a team from doing two to four months of development thanks to an insight, you're actually speeding up the process. The question is, do we understand that when we build software, versus thinking, "Oh, you're stopping me from making decisions." Stop being so arrogant.

Ryan Glasgow: It’s a false dichotomy. The best research teams we see are able to inject research at the highest impact moments in the product development process, and it doesn't have any impact on the team's velocity. It's not a decision between slowing down or doing research. The best teams are moving really fast and injecting research at the right moment. For example, for every new feature they launch with an A/B test, they're not just testing business metrics; they're also incorporating in-product surveys to measure the experience data as well. That user experience is the leading indicator for long-term business impact.

Finally, for everyone who needs to justify their work, what are the best techniques to convince skeptical management to invest properly in research?

Jonathan Widawski: It's really hard because the ROI of research often lives in the negative space—it lives in the things you didn't build, the products you didn't launch because you learned early that you shouldn't. The reality is that it's not often a straight line.

What we've seen be very successful is to start small and to celebrate the success. The best researchers we see start with the smallest project where they can drive impact. They go hard on that project, and they report broadly, trying to visualize the empathy they create with customers. They get that success, they make it a ritual, they share it, and then they repeat. That is the best way to start creating a culture that celebrates outcomes over output.

Alfonso de la Nuez: With the automation we have now, there's no excuse. I've always been a big believer that bad data is worse than no data, so let's make sure we get the right data. But having said that, today you have the tools and technology to allow for speed. There's really no excuse. I think it's more a matter of having to do some selling in this industry. We have to show them examples, we have to show them the before and after.

Discover more talks and panels from Insight Out 2025 on YouTube.

Interview questions and answers have been edited for clarity and length.