Episode 10: Frederic-Charles Petit, CEO, Toluna

Episode Overview

Frédéric-Charles Petit, Founder and CEO of Toluna, on building a 25-year-old consumer insights company for the age of AI.

Episode Highlights

• Four pillars, one platform. Toluna's model combines a global consumer panel, proprietary technology, specialist methodology, and people to deliver integrated consumer insight at scale for enterprise brands.
• Synthetic personas grounded in real data. Toluna's AI-driven synthetic personas were built on a foundation of first-party panel data and machine learning developed from 2019 — well before the generative AI boom — enabling fast, high-correlation answers to real business questions.
• AI as the engine of democratisation. Frédéric-Charles argues that AI is the most significant transformation force in market research, enabling research to move at the speed at which content is now created and decisions are now made.
• From gatekeeper to door opener. The insights function inside enterprise organisations is shifting — from controlling access to research to enabling broader teams to act on consumer data with confidence and speed.
• Adoption is maturing, not just experimenting. Customer uptake of AI-first research solutions at Toluna has already surpassed full-year revenues from the previous year, pointing to a market that is moving from pilot to embedded practice.
• Strategy and tactics in equal measure. After 25 years building Toluna, Frédéric-Charles's core leadership lesson is the need to balance long-term strategy with close operational involvement — what he calls "top down and bottom up."

About the Guest

Frédéric-Charles Petit is the Founder and CEO of Toluna, a global digital consumer insights company he has built over more than 25 years. Under his leadership, Toluna has evolved from a panel provider into a technology-driven insight platform serving enterprise brands worldwide.

The company's current focus is on embedding AI across both quantitative and qualitative research, including the development of synthetic personas and AI-powered ad testing tools. Frédéric-Charles is based in France and is a long-standing figure in the research technology industry.

Learn more about the impact of technology and AI on research, insights & analytics at Insight Platforms.

Transcript

Mike 0:16

Hello again everyone, and welcome to another edition of the Founders and Leaders series from Insight Platforms. I'm your host, Mike Stevens, and I'm joined today by Frederic Frederic, who is the founder and CEO of Toluna, one of the more well-known and established data and technology brands in the research and insights industry. Frederic, welcome to the show.

Frederic 0:40

Thank you very much, Mike, for having me today. I'm delighted to be with you.

Mike 0:45

That's great. So look, I introduced Toluna as a data and technology business, but I guess you'll probably describe it in much more explicit terms than that. Can you tell us what is today's Toluna all about? What does the business do?

Frederic 0:59

Yes, it's a, it's a great question. because Toluna has been a company that has really transformed itself since its foundation back long time ago now more than 25 years ago actually. And it's really a digital consumer insight company, that really have built a foundation, built on very strong pillars One is this global consumer panel which is a very important foundation of the group. The second part is o ur technology platform and our ability to develop technology, which is something that I'm a strong believer in. You can't create from scratch. You have to have a DNA grounded in technology. And we have very strong solutions and methodology around brand and product, and great people across the globe. When you take those four elements and then you play with those elements, that gives you what to able to do from very large mandates for, global brand, through agile quick turnaround research using our DIY platform. And now, AI based research.

Mike 2:13

So could you take us through those four components? So you've got the data. I think, a lot of people will know you for the global panel reach. I think, what is it, 70, 80 million,

Frederic 2:24

Yeah.

Mike 2:26

Can you just talk us through those four components then? So the data side of things, then the technology.

Frederic 2:31

I think it's, I don't see them as separated anymore. I see them and we see them we offer them as part of the same experience for clients because Toluna historically was what we used to call a panel company. And, Tol una is sitting on enormous first party data and one of the largest high quality consumer panels. But actually I can say like more than 80% of our revenue today, we are delivering very strong insight, to brands directly. So brands, they want solutions for their innovation sprint, their R OI on, marketing investment, the understanding of the market, where they operate, the brand, and the impact of their brand on their product. So I think this is really the integration of those four pillar that are really delivering the value of Toluna at scale globally.

Mike 3:39

So you've built the data heritage, with what you talked about with the panel, obviously the technology components that help power that. There's Toluna Start, the agile surveys component, and then you've also got qualitative capabilities on the tech side as well. Is that correct?

Frederic 3:54

Yes. We have our main platform is Toluna Start, which is a DIY platform or assisted DIY because we have members of staff that can support the client needs. And with this platform, you can do either custom research or solution based research. So you can do ad pretest, you can do brand tracking, you can do a number of different things with the platform. But we also have for some time Toluna Start Qual, which is basically a platform that helps you run asynchronous qual and it's infused by AI as well. So I think maybe we can talk about that a bit later on. But the foundation of the company are these platforms, where the platforms over time are embedding all of our solutions and become really the point of contact, in terms of aspiration, with our clients.

Mike 4:47

You've talked about AI in a couple of places, and you know that the AI is infused in the qual, that it's part of Toluna Start. A few things caught my eye recently with some of the the synthetic personas that you're using for ad testing and that kind of thing. Can you tell us a little bit more about how that's evolving and the kind of use cases that you've got specifically for AI tools?

Frederic 5:11

Yeah so I think if I may take a moment to go back in how we've built this synthetic persona, because I think this is really important to understand our legacy in, high quality first party data It's what gave us the ability to build this synthetic persona. So in 2019 my engineering team, our engineering team came to us and said, there's something we absolutely need to embark on. And he said that we absolutely need to accelerate using machine learning and AI. And I said, why? And they said Frederic, we might not know why yet, but what we know is that if we don't embark now might be too late. And so we started by using machine learning and deep learning in order to vet the quality of the panel, the quality of the response. Then we used it to do fraud detection in the panel. And then at some point we saw that every single individual in our panel was a mathematical equation. And so from there we basically understood that, or we maybe start to predict what's the best survey for each of those panel members based on, the modeling that we've done. And then in parallel we developed conversational research using AI, what we call a QProbe, where in-survey, AI asks questions that we haven't ourselves decided and asks questions with a lot of understanding of the sentiment of the respondent. Then we've used AI to build, to be able to code sentiment at very large scale, coming back to qual, and on that foundation we build the persona. But the persona is not taking Frederic, which is a panel member, and building Freddo as a synthetic persona. The persona takes a few attributes, and then the rest is being built by the system. But our personas are really grounded in truth. They're grounded in the Toluna panel, which is one of the, I think one of the best panel worldwide, and with the historical understanding of what respondents are doing. And so you have to think about it as and I'm not using that for a buzzword, it's true it's a big agentic system that is basically described as a synthetic persona.

Mike 7:56

Okay. Fascinating. So the origins of this are actually in machine learning and predictive analytics before this whole Generative AI boom. It's not LLM-based.

Frederic 8:06

Yes, Generative AI came in, but we had this very strong foundation of developing AI tools. Then what we did, if I may say what we did, we said we could do so many of these personas, but why don't we start by answering key business questions? Claim testing. So we started by claims, and then we said maybe we should answer a question around ad pretest. And so we've, this is the new Toluna, which is not just using this persona for field work. But using this persona to answer, at fast turnaround, and with huge correlation, business questions that our clients are asking every day.

Mike 8:53

Can you talk a little bit about how your customers global enterprise clients and insights teams work with Toluna for, ad testing, pack testing, using traditional methods. But how are they reacting to and adopting the synthetic data? Are there different segments of customers? Some of whom are very wary, others who are very keen adopters? How is it going down?

Frederic 9:20

I don't know if it's a good news or not, but in a way, Toluna has been there before. When Toluna launched online research, we got a variety of behavior within our customers. There were some customers who were cautious about it and they won't experiment, and they're not sure. There are customers who are very progressive about it, and they want to go all in. I call them the'all-in' customers. And the difference is that Toluna has 25 years of legacy of being a consumer insight company with very strong first party data. So it reassures our customer that Toluna is already providing them with either strong, high quality data or insight, and that Toluna knows what it's doing when we are launching these new products in market. So I think that there is a gradation of usage, but I do believe that we are reaching a point of I don't know if we're at the peak, because we only know after the peak when we are at the peak but we are at the peak of piloting with customers.

Mike 10:27

Okay, so it's starting to mature and actually, land in their business process more than just an experiment now, is that what you're saying?

Frederic 10:34

Yeah actually it's funny'cause I looked at the I'm not gonna tell you the number but I look at the revenue that we did full last year with AI-first solutions, AI-infused solutions, qual and quant, and I looked at this year(June to December) and we're already way above what we've done last year, in qual and quant. So I think that the market is still in its early hours and it's a question of days.

Mike 11:09

Okay, so for those early adopters, are they maturing in the way that they're using maybe synthetic as part of a bigger process? Does it mean that they can test more things earlier and then still use human feedback? Are they substituting in some cases or is it more about complimentary tool?

Frederic 11:28

That's a very good question. I think that if you look at AI, what it does the usage of AI, it's making an explosion of content, an explosion of variation of content. And if use the old research methodology, you're never gonna be capable of testing all those content, you're never gonna be able to deliver research and insight at the speed of AI if you're not using AI. So I think this is really one element is that, in my view, AI is really the largest, biggest transformation force in the market research industry. It doesn't mean that you don't need strong data, high quality data, high quality respondents; you need them, but in a different way that they can help you over time, build strong AI capabilities. And so we say at Toluna that we're bringing research beyond limits, but the real aspiration is making it limitless. And so that's really our aspiration to democratize research. It has been our obsession for 25 years. How do we democratize research? I think online did some democratization, but AI is really the engine of democratization within the enterprise, but also maybe other sectors.

Mike 12:56

So if I just paraphrase, if we've got generative AI creating all of these different variants content, campaigns, packaging, claims, whatever it is they need to be validated or proven. Having the synthetic capabilities with AI research is gonna make teams able to better cope with that kind of volume. Is that correct?

Frederic 13:19

It's one of the aspects that would make them be able to test at scale. The second aspect is that right now research is I don't know if you say that in English confined to certain territory within the enterprise. AI, and if you build strong AI packages, would give true democratization of research within the enterprise. More people would be able to use it with more confidence. So I think it's, yes, it's one of the aspects, AI and synthetic persona, but not just synthetic persona, what we call also direct to prediction products will do that. For instance, if you look at ACT Instant AI, which is our ad pretest, you just have to pull your ad in the system, and then it creates the questionnaire, it interrogates the 200 or more personas, and then you get the result in, in, in almost no time. And so you you can do that to a variety of formats, you could very quickly build benchmarks. So it is creating a game changer in terms of the quality of the information on which you are basing your decision. Under one condition: the quality the AI, the quality of the synthetic persona, it's the same problem, it's the same challenge that you have with any other source of collecting insight, whether it's telephone or face-to-face or online, and not AI. So the quality of the persona is key.

Mike 14:56

Okay. And that's powered by human data, the kind of quality of the human data that goes into it.

Frederic 15:02

It's powered by first party data, but not only that, there other systems around it. It's fascinating to be honest. When you look at it from inside.

Mike 15:14

Yeah. Okay, great. What if we just look beyond obviously the scope of what Toluna is doing is quite broad, you've got the data, the solutions, the people and the technology but what more broadly is exciting you in the insights industry? What are some of the big trends that you see? What changes do you see happening across the research and insights space?

Frederic 15:38

So I think that first, I think the insight industry at a time where truth is a real challenge to get I think that the industry has a great opportunity'cause we, we are industry. We are people who are, we want truth to be told. And that's very important. Like the response to the survey and the research we're doing is used for making very important decisions. So I think that when you look at it from that standpoint, think that it's very exciting for the consumer Insights Industry. So I think what's interesting is that we live in we live in a society where truth is hard to get, including, but not only on social media and the insight industry with true first party data grounded in truth, gives you that ability to get the truth about my product as far as Toluna is concerned the product, the services, the quality, the feedback. All these things that are critical for our clients and I think this is it's a great place to be as an industry and I think it's a bit underestimated by our industry that we have this we have this very significant position that we have very strong first party data and first party data grounded in truth is critical. I think that the revolution with AI is gonna solve some of the fundamental. Historical that the industry had, which is how to make this industry a mass market. And, that's really if anyone could use research. Now, obviously Una is very focused in the enterprise market. So we're very focusing into client who have global or local or regional product and brand and to develop and grow. But I think that even the democratization within the enterprise, has been somehow. Delivered, but not fully by online. Somehow delivered, but not fully by DIY. I think that there is a huge opportunity to make it scalable and require a complete change of, system and processors, a foundation of the way we worked, but ultimately what we deliver is the same. It's this ability to give an insight that would help our client to have a better product earlier on than later on and to be able to read the sign in the market that becomes signal that become trends. And I think this is what's very exciting. If we take it like that. It's very exciting. If you take the other, take the other and we're between friends here, if you take the other flip of the coin, and you say all the changes that need to happen, then it might be a bit scary.

Mike 18:30

Yep. So that's the flip side that I was thinking about, which is, we have the potential to democratize to get more people, even in enterprise businesses, more hands-on with data and insights and research. But the structures have historically been small teams through which that's funneled. So a team will manage external partners and then manage internal stakeholders. Do you see that being disrupted over the next little while? Do you see changes to that structure?

Frederic 19:01

I think I think first the way the research company is gonna be disrupted, the consumer insight company, that's the first things that are gonna be disrupted because if you want really you really want to scale AI and deliver AI, you need to change the way you are working. And so that requires significant change. So that's point one. The second point, yes, I think that is actually a change over time of the insight function, but I'm a big believer that's a reinforcement of that function, provided that function doesn't become like the legacy stage gate for research, but rather, the door opener to the organization for knowing the pulse of consumer at the speed of AI, being all the things we said with the on-demand economy. We thought that DIY would open up research at the speed of the on-demand economy. I think AI is ultimately the driver of that. And so if you think about your function in insight, so the research function, from, deciding who's doing the research, which partner to go, and I think that the research function more and more, and I see that with client, I see this openness with our clients of understanding that from being a stage gate, they become door opener. And I think this is really what the research function and the insights function within the enterprise has always been a door opener, but now it's opening with quality, with ground truth. I don't know if you said the floodgate of, democratization within the enterprise, how can you give it to someone who is not a researcher? That they can make decision, or use that as a decision marker in their respective funnel.

Mike 20:55

Okay. I like that from a gatekeeper to a door opener. I think it's a different kind of framing of it. Yeah. Very good. Okay obviously, lots of change structurally. You talked about democratization within the enterprise. Being able to get more people who are not researchers able to use research and insights in their planning. What do you think, there's other angles to democratization, which is small, medium enterprises being able to use it. More startups, like MailChimp did for email marketing, or Canva has done for design. Do you see opportunities there as well, or are you primarily focused on the sort of enterprise?

Frederic 21:34

Toluna, historically has been focused into enterprise, and and we're very focused into the democratization of research within the enterprise market or and within our Toluna within our research and agency market. So that's where we focus, but clearly. I say all that and there must be something where democratization doesn't stop at the size of the company, doesn't stop at it's really how can we make a good product that, I don't, I, I never thought about one size fits all strategy because I think clients are different, vertical are different, their focus is different. The level insight they need is different. But I do believe that there is, yes. The market can open that and yes, that's why I'm talking about ultimately the mission to make this market a mass market.

Mike 22:27

Great. Just changing tack a little bit. You mentioned 25 years plus, founding in the Toluna business. Can you talk a little bit about some of the lessons you've acquired along that journey? Maybe some of the important things that, you know, that you've picked up as a founder, as an entrepreneur, what have been some of the biggest lessons that you've learned?

Frederic 22:49

I think there are many lessons across the years and I think that attention to detail is always very important. Being really attentive to detail, going in stages. So I always say we can have the ambition of going at the top of the Everest or the Himalayas, I used to say when we started, let's go to the first floor first and that doesn't impede you. So having balancing between strategy and tactics, and especially in market like that, which are quite challenging, like staying focused on your on your strategy, being able to, to adapt to the market, AI, tariff economic headwinds, global issues... all that. So I think being always balancing between the two. And trying to really articulatte your strategy in a way that, that balances between those two pillars of strategy versus tactic. I think that's critical for any business. If you do just one if you just do tactics you might not be successful long term. If you just do strategy you might have issues along the line.

Mike 23:56

I heard it expressed once as as the need to do the 30,000 foot view and the three foot view at the same time, you need to

Frederic 24:02

Yeah. So I will tell

Mike 24:03

jump between.

Frederic 24:04

which is about founders, because I think very strongly about it after all these years. that the idea that a CEO stays on his vertical, like we say in France, the'la verticale du pouvoir', the vertical of power, and he is sitting in the cloud over there and not going and having the ability to deep dive into a subject. This is contrary to what I believe. A CEO and a founder, especially in the new world in which we live, should be able to inject himself or immerse himself with the team. And the priorities are, what are the priority this week? I mean, and what are the priority next week, especially given the uncertainty in the market. And so I'm always very worried when people are calling to delegation, delegation of power, as in inserting yourself in some of these decision processes. On the back of a hierarchy or an organization framework, all that, things are so flexible in today's world that I do believe that I wanna be able to, look at the GUI of Toluna Start one week and I think that my point here is the ability to being able to operate, at he helicopter view as we say, but also going to much detail and that ability, I think is required from CEOs. And founders are used to doing that. So I think it's extremely important, in today's world.

Mike 25:47

Okay great insights. Don't be don't expect that you're just gonna sit in a corner office and and delegate to everyone, get hands on, stay hands on.

Frederic 25:55

And always be careful of the people. Always be worried of the people that tell you that you should not participate in that decision because you're the CEO. It's not for you. You should be worried. Now that doesn't mean that you should inject yourself in all the decisions, and you should be able to look at what makes sense for you to get involved in. Is it really critical? And so that's basically what I think. Is is required of CEOs in this world.

Mike 26:25

Fabulous tip, I think for anyone else founding a business in the research and insights space, anyone founding or leading a business to to end this conversation on. Make sure that you stay involved sufficiently in the detail as well as keeping an eye on the strategic direction of the business.

Frederic 26:42

Top down, and bottom up.

Mike 26:45

There we go. Wonderful. Thank you very much for your insights, for your time. You've been very generous. I know you are, you're incredibly busy. I do appreciate you spending the time talking to us today, and we thank you very much. This is this is another episode of the Founders and Leaders Series. We have many more in the pipeline and I hope you'll be back to listen to more. But for today, that's all we have from Frederic-Charles Petit, Founder and CEO of Toluna. Thank you very much.

Frederic 27:12

Thanks, Mike. Very happy to be with you. Thanks. Bye.