From Panel to Platform: How Toluna Is Redefining Research with Synthetic Data and AI Agents

In this episode of the CEO Series, Leonard Murphy sits down with Frederic-Charles Petit, CEO and Founder of Toluna, and Renee Smith, Toluna’s Global Head of Innovation, for a wide-ranging conversation on the future of insights. From their early investments in AI and machine learning to the launch of synthetic sample solutions, Toluna is leading the charge in transforming how data is collected, interpreted, and delivered.

The conversation dives deep into how synthetic personas, agent-based systems, and quality-first data strategies are driving innovation—while also challenging outdated models of research, project work, and pricing. Whether you're curious about where the industry is headed or how mid-size firms are gaining competitive edge, this episode offers a rare behind-the-scenes look at the future of research in the AI age.

Transcript

Leonard Murphy: Hello everybody. It's Lenny Murphy with another edition of the CEO series of interviews guests. Usually, it's only kind of a one-on-one, but today we've got double the trouble. So, I am joined by Frederri from Toluna and…

Frederic-Charles Petit: Thanks. Thanks very much, Lenny.

Leonard Murphy: Renee from Tuluna. Thank you guys both for being here. it's wonderful to have you both joining the show.

Renee Smith: Thanks for having us.

Leonard Murphy: It's so for people who've been living under a rock, maybe just do a quick bio introduction for both of you, but then I know we have a much cooler topic to get into than just the bio. So, Renee, ladies first, we'll start with you.

Renee Smith: Thank you so much. Renee Smith, I may have seen many of you at conferences at various times. I became part of Toluna a couple years ago when Toluna acquired Gutch Check and I'm the global head of innovation at Toluna. Happy to be here.

Leonard Murphy: So, since you all say it, we met when you were what chief research officer at Canar if I recall correctly.

Renee Smith: Yeah. Thanks, Lenny.

Leonard Murphy: So a long and storied career in thinking about cool stuff in the industry. You're welcome Frederick.

Frederic-Charles Petit: Hi, I'm working from home as you see and I have some kids coming back from school. So you would probably excuse me for that. So yeah, I I'm the chief executive and the founder of Toluna. and yeah, other than that, I'm very excited to be on this call and really the fact that we're both on the call is reflective of the transformation we're going through and the transformation our industry is going through.

Leonard Murphy: Yeah. No, all I'll tee it up for the listeners. first a little bit of history. obviously been a big fan of both, Frederri and Renee as individuals as well as Toluna for a long time and have watched the transformation of the company from a panel provider right back although Frederick you're probably saying wait we were far more than that and yes I would say that you were but that was kind of the perception into now a really integrated Restec platform with sample and variety of solutions. What I always liked about you, even though you were a panel provider, was that you had a deeper engagement with respondents and that you had a lot of data on them.  And Frederick, I even remember you doing a presentation at SMR. And I always bring this up when we talk that you were envisioning the future where you were leveraging a very comprehensive data set of the consumers to drive more individualized experience. and you just were more than just slaying sample. now we fast forward to a couple weeks ago you guys had announced the launch of your synthetic sample solution and that prompted wanted to have this conversation because I know that synthetic sample in the way it's being utilized today was something that you guys were thinking about years ago in terms of its application. We may not have called it that, But this vision of how to utilize existing data to drive more value in the insights process and to transform that to move away from the bespoke project into something that looks a whole lot more organic over time was before AI was cool from that standpoint, here we are that the technology allows us to do that in a much more efficient way. AI's unlocked capabilities to bring that vision to life and you're out of the gate with a solution around that. So for this call I wanted to dive into that with you guys and for our listeners to kind of get the perspective of one of the major players in the world on this transformation from collection into data utilization and how that you're powering that. Again Renee I'm going to go to you first because of your background and the chief research officer right you've been thinking about these types of things for a long time so kind of give me your take on this transformation and…

Renee Smith: Okay.

Leonard Murphy: Where we are and what you're seeing happen Now,

Renee Smith: Yeah. I think that one of the things we've seen just consistently over the 25 years I've been in market research is the not every decision, but the speed at which some decisions have to be made by our clients.  And again, it's not every decision, but there are some decisions that benefit from being able to be made in a series of, hours or very quickly. And so when I think about the uses of data, I'm always thinking about to what value. So you could create synthetic data for the sake of creating it I guess but what are the use cases where it can facilitate more speed with a decent level of accuracy? What are the use cases where it can facilitate better measurement not even for speed but just better measurement in general.  So I've always had a pretty kind of eclectic view of data and for me it's always about what is the purpose and so I think that the fact that we now have synthetic options in our toolbox as well will over time be able to say here's the right use case for synthetic here's the right use case for hybrid here's the right use case for the human data and just before I throw it over to Frederri I do want to say that in terms of data utilization it's important that data has sufficient quality. So just for the sake of using data it has sufficient quality and the industry you probably remember two and three years ago especially was getting hit by a lot of fraud and bots and stuff and…

Leonard Murphy: Still is. What are you talking about? Two or three years ago. Yeah, a agreed.

Renee Smith: What I will say though is that some companies Toluna is among them have made some good strides in detecting that and if we hadn't it'd be a very different story in terms of our ability to actually use the data so good  Human data goes hand-in-hand with good synthetic

Leonard Murphy: And Frederick, I want to get your perspective on this as well. I just want to reiterate what you said. the, …

Frederic-Charles Petit: Yeah.  Thanks.

Leonard Murphy: The fraud issue when we're dealing with one project was bad enough, but with the advent of AI and synthesized data across massive data sets, it's peeing in the right? I mean sorry just to be somewhat crude but it's just a contaminating factor of the risk of garbage out scenario in the era of AI especially as we move towards an AI first approach for some business issues it's just massive so now you've been visionary I'll say it for you kind of thinking about where the industry was going to and you're also tasked with driving the financial performance of a large global business with investors etc etc. So, what have you been thinking about this transformation? How do you see playing out at kind of that macro level within the business itself?

Frederic-Charles Petit: So I think that first it's really I want to chime in on… what Rene says. for us it's not we want to build synthetic persona or AI just for building it because we're not at a size where I can just decide and do that anymore in a way which is a benefit and a constraint. I mean it's not a startup and so the first there's a couple of element that we need to take into account when we innovate and I think the first one is there any benefit for our clients are we in what's the clear benefit for the client what is the accuracy of the data or insight that we will be transferring to the client. And I say that because a client might agree that a startup with low fidelity on its new product the client might accept that because it's a trial and they're not making any decision on that. With Tuluna, it's very different because our clients are making very important or very frequent decision based on the data or the insight that we're providing them.  So that's the second part and the third part is I'm always a big believer that I'd rather bring the innovation inside than get the innovation from outside. By that I mean that I don't think that Tuluna or any player in the marketplace. I've seen it with online can stop the emergence of new technology and so therefore I'll rather be part of embracing it early in advance rather than being subjected to it or by it I don't know if you say that in English later on.  Now I would say that with a caveat is that I've also observed in the market that is there really a benefit of being the first mover. I don't know but that's what Toluna does.  I mean that culturally to Luna like to be first in not all the time and so we are in this position where we can launch this new product and solution and that's not a panel by that I mean I want to be very clear to not launching a synthetic panel and I explain that maybe in a moment we've been in this position because we took financial decision which might  have been wrong and turned out to be right back in which seems to be a long time ago. In 2019, I had a couple of guys, people coming to me from the engineering team. It came from engineering saying, "We want to invest in machine learning and AI." And I said, "What for?" And they said, " let us show you what it does, but we don't know what would be the result of that." And you always tend to do best as a company. So what we knew best was at the time was we knew very well respondent. We knew the problem with the respondent. We knew the benefit with respondent. We knew the trick with the respondent. we knew the fatigue sometimes with respondent. So we basically use that to build a number of machine learning model in order to identify that and so we built the foundation and very clearly over time and iteratively I looked at this as being the foundation of or…

Leonard Murphy: Mhm.

Frederic-Charles Petit: The building blocks of the AI strategy for the group.  So I don't know if I'm clear enough in my response but I do believe that these are decision that you make long time ago and it's online.  I started Toluna and it took to five years. it took for Toluna to start seeing to being in a position where we knew that we will go out of the nuclear winter of the new economy we said at the time took five years. So you have to be patient with this decision and you have to prepare yourself and at some point there is a tipping point in the market where it grows and I used to say I think 2 it took 10 years from offline to online data 5 years from online data collection to DIY and I believe it's going to take two years from DIY to AI for the research industry. So that's my big and so how do we do as CEO manager leader Rene on innovation to make sure that we're relevant in the market that we bring this value and that we bring a value that satisfies all the stakeholder that's really the

Leonard Murphy: Mhm. No... I think that's fantastic. so I describe this as an arms race,…

Frederic-Charles Petit: You're right.

Leonard Murphy: On two fronts. First, there's still the amazing week changes just in the capabilities of AI of LLMs, right? that is still a very unsettled piece of technology, meaning you start building something on chat GPT1 and guess what, right? Here comes DeepSeek or Grock or whatever that just blows it out of the water and the next week it'll be something else so there's just this continual massive shift in the capabilities of the technology itself that's leading the adoption curve particularly for B2B so we have that and then we have obviously the companies that are trying to take these pieces and build solutions with product market fit happening and I think what we've seen my read is that buyers big brands right obviously for a variety of reasons have been waiting to see if we reach some level of par and that's not happening right there is no wait until something reaches maturity we're not in that cycle I would actually argue from the standpoint of maturity we bypassed that a long time ago. Now it's just incremental innovation over and over and over again at a very rapid scale. So there's no bystanders now, Maybe they were trying to do that, but now it's like, crap, we have to do this now. We have to embrace these technologies before they really start limiting their ability to create more value within the insights function.  So that's my long-winded way of teeing up to say what are you seeing from that standpoint right you made these investments Frederri I agree that the life cycle is very small now from an adoption standpoint what are you seeing in terms of client adoption are you seeing clients wanting to embrace these technologies maybe still cautiously are they going allin are. Hey, just what's your take?

Frederic-Charles Petit: Renee, maybe Rene first and then I'll do it.

Renee Smith: So we have an AI client advisory board and…

Leonard Murphy: Okay. Mhm.

Renee Smith: We've had it for about a little over a year and we see many levels of kind of readiness and some of it has to do with initially just kind of getting their arms around in the insights area, but some of it also has to do in very regular  related industries, finance, healthcare, there's just a lot of infrastructure that sort of has to be, put in place. So, there are going to be pockets that are not going to be able to lean in and adopt quite as fast. What I've been very pleased about as we've taken out rapid claims testing with harmonized personas. So, that's a quantitative solution for claim screening. What's very interesting is the number of clients that first of all people want to try it.

Leonard Murphy: What's this? Yeah.

Renee Smith: They absolutely want to try it but the number of clients who have responded because they're starting to think about I could use something like that to really do agile iterations.  So, we've been talking about innovation processes at brands becoming more agile, and some had been able to implement it, but I think others now wait, I actually can potentially implement it. and so I won't be surprised over time if actually what you start to see is not necessarily even an RFP that comes out to be the agile provider but that comes out to be the agile screening provider and the agile concept testing provider right it at that very early stage so there's definitely some demand there and I would say that the other difference that I see and it's kind of related to the comment that you made Lenny but compared to in 2000 and…

Leonard Murphy: Okay. Right.

Renee Smith: The adoption of online. at that point the insights industry seemed to have a culture of let's wait for one or the two of these big brands to do it first and there's no waiting now, Everybody, as you said, sort of realizes they should get in the game. So...

Frederic-Charles Petit: Yeah, I think we came to market saying a AI everywhere at to do now and that's true I mean because of the things that I've just described so our clients have been interacting with AI for some time the way we procure sample was driven by machine learning I mean it's not digital twins by no mean we're not fulfilling the survey without telling the client that it would be digital twins then all the quality what we call QPER which is a multi-layer AI and…

Leonard Murphy: Excuse me.

Frederic-Charles Petit: Traditional methodology quality rather than just fraud detection because  something that there is only pod as a problem quality that has been really there then using some of that work that we've done around quality that's when we introduced Q probe I'm giving you the name not to advertise them but just to say Q probe being probing live in the questionnaire through AI and I don't know what AI is going to probe  It's like so we started doing that and we saw that it was adding 40% of good quality open end and then we launched because we did a lot of detection of problematic not to say fraudulent open end with machine learning and AI. We kind of leverage that technology that we built to do coding sentiment analysis and so clients are playing with that every day they go on the platform they can decide and we leave them the choice to decide and we always tell them that's driven by AI so we want to make sure that client understand what we're doing it's and so they've been playing for a long time when we decided and actually we decided to go for building a synthetic panel for a better word. I mean in the US initially the idea was we want to make sure that it's not just a capability we want to make sure that it reflects what to do has become in this market. So we want to make sure that it can be used for use case by client because only by testing use case they would adopt and then we can launch more use case.  So I see a good adoption. I think that it's a question of making sure that we're very transparent with client with the benefit and sometimes there is not only benefit in any products like online there was no just benefit in using online there might have been some downside telephone versus face to face at the time. And I think the position in which Tuluna is great because Tuluna has changed so much that we're kind of agnostic. people say do you believe in the future of online panel and I said what I'm going to give you the answer but first of all I want to tell you I'm agnostic because of what the company has how the company has transformed.  So this is really the strategy of the company is making sure that client can use at a number of different touch point with the company so this client when we bring a breakthrough rapid claim test is a breakthrough I do believe it then they're not surprised they understand that the company has invested in it that we know what we're doing and that we're not just trying to sell one product because we have a broader responsibility in the engagement with clients. So we can't make it wrong.

Leonard Murphy: Yeah, sure.

Frederic-Charles Petit: We can't sell it and then go for a run. And I don't know if it's clear enough, but

Leonard Murphy: It is. tell me if you guys agree with this is how I see the future kind of shaking out, and I think it's also a description of probably Tuluna. so fundamentally the research industry was based on monetizing process and outputs. That was how we made our money. and those are the two things that are impacted fundamentally by from an efficiency standpoint. but in order for that to happen, AI has to have access to data. So I switch back and think that okay so fundamentally what the research industry really has as an asset is data as an asset and that data is the connectivity to humans right that is the fundamental difference between chat GPT that's pulling in from everything else and that's great but it's not quality data that is predictive of human behavior to make business decisions off of would be my argument, To do that, we still need something that looks like research from the standpoint of collecting information, from a sample that is structured in a way that we can predict outcomes.

Frederic-Charles Petit: Yeah.

Leonard Murphy: You can make business decisions off of that. But is It's no long process of collecting the data. and then fundamentally we get to a point where you have data and then you have making sense of data. which is kind of the consulting piece of things now kind of the implications and outcomes. so the companies that have access to data and have embedded the technologies to be able to extract that to pass it on to people to make sense of it. Those are the businesses that will create real value in the future. That's my take and go ahead. Okay.

Renee Smith: I would have one oops I would have one nuance to that which is and I'll use the rapid claims as an example is that synthetic personas our system the way it's set up is actually not asking a large language model to pretend to have a profile which is what was kind of the first stage of things feed some characteristics in and ask the large language model to be the persona okay so be Shakespeare be consumer ABC what we actually built and I think that's one of the reasons why some of the persona work started out in qualitative.

Leonard Murphy: Mhm.

Renee Smith: But if you actually think about quantitative and what we're doing, we actually knew that we wanted to create samples. And so we needed the kind of variation that was going to reflect a sample. We couldn't just have segments that were giving us the common answer for that segment basically. So the first thing I want to say that I actually think it's the individual level data that is particularly important as the seed cases. so can other data be helpful? Yes.  But I think the individual level data and then in terms of what we actually built, because we don't use the large language model as the persona, we actually built an agentic system of about 20 agents that are doing you might say that they're doing element some of them are doing elements of the human workflow. So there's a agent that programs it in prompts and so forth in the way that a persona can actually digest the information. So the reason I'm making that nuance is that I think that right now there's a lot of talk about things like agents and being using them for efficiency and stuff, But you do actually have a pro have to have a bit of a process to use personas and…

Leonard Murphy: Okay.

Renee Smith: being smart about what that process is going to be is actually also going to be IP. so it'll be data but it will also be the way in which the systems are actually configured that will also I think bring value. And so when we talk with clients now, we have one slide where, we try to make it as least complicated as possible, but we actually talk about, hey, you're not just plugging information into an LLM. There's actually this kind of set of checks and balances and these agents that are challenging the answer and all that sort of thing. So, I think that there's other IP there, I guess, is all I'm saying, in addition to the data.

Leonard Murphy: I think that that's a great point. And fundamentally I am a consultant, right? So, the only thing I have to sell is my brain. but yet cannot it's foolish. I fought engaging with AI for a long time because I thought it was going to devalue me, and recently really the last 69 months I got over that and discovered as a superpower and with that very same concept of look the IP is still in my head the process and how I can configure those processes of utilizing LLMs effectively to take what's in my head and make it deliverable that's IP as well. So that's wonderful that you point that out. I think everybody needs to recognize that it's very complex and just the chat GPT perplexity, that's a different thing. that's tip of the spear surface level. that's not really where we are from a transformational technology. and how it's changing things. All right.  So we could talk for a long time and I want to be conscious of your time as well as the listeners. when you look at the competitive landscape obviously most of your core competitors that you would think of particularly those that play in the sample space as well as sample plus data collection. Everyone seems to be scrambling to move towards a subscription-based model with the assumption that their data becomes a feed for some LLMs whether they've built those internally like you have or externally and that's a big difference for how the industry functions although when you started transforming into a Restec company you launched that subscription type of model at that time so  you were kind of thinking ahead in that respect as well. So what is the business model of research look like as we move forward? is it clients buying a subscription to a suite of services which includes access to data and they're paying a flat amount based on volume of usage? And then they may also execute bespoke projects in that agile approach that you were talking about, Renee, to fill in gaps of information or explore things deeply. I guess to sum that up, do you think we're moving away from the project and even the SAS license, I'm not sure that's really the right model going forward either, to something different that is more of an ongoing subscription that looks probably similar to what people are paying now for use of LLMs. What do you think?

Frederic-Charles Petit: Yeah, I can Rene I mean I can go on that. I think it depends…

Renee Smith: Mhm.

Frederic-Charles Petit: Where you are in the value chain. And by value chain, I don't mean that there's better than worse in that value chain. But if you're a pure panel player, a panel company or a marketplace or I think that the model is going to be different than if you're a company that's whose mission is to generate Tuluna today trailblazing insight for client. I think the model is going to be definitely different. So for the later I think that we've done a subscription model some time ago and at the time the subscript it was quite new

Leonard Murphy: It's sassy. Sorry.

Frederic-Charles Petit: Because I look at it and I said it's almost like it's not really SAS but it's SASish I would call it this way in a way. I think we're committed to I don't know how you call that sassish. and think I don't think AI all would replace fully in our industry at least not in the next few years. The need for people and I'm talking about researcher or people managing that infrastructure I think that the industry is take enabled you could and it's not even fully DIY that unless you're quantri maybe there is a form of glass ceiling for pure DIY player  in this industry and that's a recognition that Tuluna took on his journey to notably M&A. So I don't have the silver bullet for how the model is going to evolve. I think that you've got a good one with the way tokens are playing with the LLMs and all that. I think that's a good understanding of what's the unique price. Okay, maybe. But for people like us who are delivering more just than a prompt and we're delivering more than just a we're delivering solution which gives you the ability to measure the equity of your brand or the efficiency of your investment in ad or the new innovation in your pack.  I think that we would need to take into account those dimension into the pricing. I do believe that facilitating the interaction of client with those technology and services through a subscription is going to continue and this is something that clients are going to be adopting more. but it's going to take time. So I don't have the silver bullet. Be honest what the model would be I think it all comes to what type of value do you deliver if you deliver a complete by that either if it's synthetic or a real individual of complete then it's going to be a different model than…

Leonard Murphy: Right. Rene,…

Frederic-Charles Petit: If you delivering insight and exactly the same way that the industry is working today so I think that we will  find a good balance in terms of the business model moving forward. But that's going to be iterative in nature.

Leonard Murphy: what do you think? Especially since you've come from kind of the service side of the industry into the tech side over the past few years. what's that? Yeah. Yeah.

Renee Smith: We had subscriptions at Gutch Check. even though it was full service and part of the reason for the subscriptions was and I wouldn't call it it wasn't SAS at all because it wasn't technology enabled particularly but part of the reason for it was just the idea of make reducing friction for clients to be able to get a project off and running basically. So, I think we have to also think about that sort of element in terms of the business model. It has to be Easy for them to buy. I don't know. I'm just going to speculate here,…

Frederic-Charles Petit: Yeah.

Renee Smith: But we may possibly start to see a difference between And what I mean by that is that when we look at the future, we know that there'll be AI infused traditional methods will continue for a while as Frederick was talking about,…

Leonard Murphy: Yep. Open.

Renee Smith: But we also see the conversational path and we also see direct to prediction path and the direct to prediction might be more amendable for go decisions and the conversational might be more amendable for kind of the why around it. And so it's possible that maybe there's subscriptions for the number of predictions you get out of a direct to prediction system but then there's the insights be where the humans are putting more analysis around the conversational so I don't totally know…

Leonard Murphy: I agree.

Renee Smith: But I think that if there are differences we might see different dimensions on which that plays out as well. Yeah.

Leonard Murphy: And I think one thing clients will have to adjust to is that if we're driving with an ongoing base of human data, incentives still are involved, we're just sucking it in off of everything posted online. and…

Frederic-Charles Petit: Yeah.

Leonard Murphy: Hopefully that will not be a challenge. it's kind of like when social media analytics emerged. it's going to replace research and it's cheap. No, that never happened  and clients recognize there is a difference between the type of data you got from social media analytics versus research and they'll still recognize that the value I've been playing this idea in my head even as a consultant of value delivery based pricing to shift away from kind of the commoditization component and I think AI does have the potential to commoditize all types of stuff so particularly information if we don't handle the education component of that correctly that the value delivery is very different than just sucking up a whole bunch of stuff and going on chat GPT and ask for something that we cannot verify whether it is actually actionable or not. 

Renee Smith: And I want to emphasize too that the part of the value so if you would have asked me because internally I was asked this last September when we were talking about we're starting to work on the positioning for…

Leonard Murphy: All right. Right.

Renee Smith: What became rapid claims with harmonized personas. it was much more at that point a sample play. I'm just telling you how the positioning evolved and once we started doing all of the testing we collaborate with the engineers to and great R&D team and engineering and seeing what could be built.

Frederic-Charles Petit: Yeah.

Renee Smith: I'm actually of the opinion that there's going to be more value that's going to come out of measurement because we need to think about these personas. If we only think of them as a sample source, we will lose the value. we will miss out on what we can bring to clients because we want to think about it as what can they do that a human cannot do and so in rapid claims as an example they don't get fatigued so they can have more claims that get evaluated now that might seem small…

Leonard Murphy: Absolutely.

Renee Smith: but we've got lots of other ideas about what they can do I'm not going to speak about them but what they can do for measurement so if we start really seeing that we can truly measure things that couldn't be measured before. That has a big value.

Leonard Murphy: Yeah, I think that goes back…

Frederic-Charles Petit: Lenny, you're frozen. Lenny, I just want to say that you're frozen on the image, but that's just for the viewer later on, maybe. Okay.

Leonard Murphy: Renee, am I frozen to you?

Renee Smith: No. Okay.

Leonard Murphy: No. Okay. Sorry about that.

Frederic-Charles Petit: Just to me. Okay.

Leonard Murphy: That may be the gods of the internet saying we should probably wrap this up. So, God, this is a great conversation and I wanted to have this for our listeners to get kind of a glimpse behind the...

Frederic-Charles Petit: Okay. Yeah.

Leonard Murphy: behind the veil, if you will, of, leaders in the industry, how they're trying to think about all these things because it impacts everybody obviously because everybody's having a variation of this conversation somewhere in their organization. As usual, you guys have been ahead of the curve and I think that's wonderful and that's just one of the great roles of Tuluna in the industry of Okay.

Renee Smith: I want to throw out something else too, Lenny, that you haven't touched on, but how does size of the company matter? so I was at Gutch Check. I went from Canar to gut check because I wanted to get in a place where maybe things could move more rapidly but that was a $20 million business and so if I'd been there and AI hit wouldn't have had the resources right I'm not saying Toluna is the only company but I actually feel very happy to be at Tuluna because it's the midsize right and…

Leonard Murphy: Mhm.

Renee Smith: We're starting to punch up there's more resources but there's also still more scrappiness. It's not perfect.

Leonard Murphy: I agree with you 100% and

Renee Smith: There's not as much scrappiness as Gutchuk.  So I also think that it's possible that mid-level players are going to be, five years from now, they're going to be in a very different spot than they are today potentially because they had the minimal amount or a decent level of resources to play this game and less solidified infrastructure or still that more scrappy attitude or whatever. the

Leonard Murphy: the large players, God bless them. They're doing what they can do, but they're fighting a defensive rear action as well, the threat of cannibalization for the large global players is significant. and to your point, not only is there kind of the inertia component of the big companies, there's also real substantive challenges in protecting revenue that has dramatic impact, especially if they're a public company, on profitability, etc., etc.

Renee Smith: So the grit report, the grit top 50 might look very different in five years than Yeah. Yeah.

Leonard Murphy: I would agree. I mean as we're having this conversation Canar is going to go away and at least in its current incarnation. I would also just point out see if you guys agree with me. It was interesting that announcement of the Canar breakup which we all kind of saw coming. the real value there was numerator and royal panel right that's the big multi-billion dollar transaction they're looking for.

Frederic-Charles Petit: Okay.

Leonard Murphy: It wasn't the rest of Canar. So, it was the data asset. and I think that people who pay attention should recognize that that was a pretty significant signal for the industry. it wasn't Miller Brown, it wasn't TNS, it wasn't, the old all of those pieces.

Frederic-Charles Petit: Right. Yeah,…

Leonard Murphy: It was the data asset. So, which you guys I would say will be in that league as well at some point in the Yeah.

Frederic-Charles Petit: We are.

Leonard Murphy: Yeah. All right. Anything else you want to mention that we didn't get into?

Frederic-Charles Petit: No, I would say it's to be continued…

Leonard Murphy: Final thoughts.

Frederic-Charles Petit: Because I said to my team we just started the race and it's just I think that you talked about an arm race earlier on. I just call it a race and so we're in the race and we have to drive that vehicle that's the difference as we build it.  So that's what's amazing about it.

Leonard Murphy: I'm not going to bet against you in the race. So, that's great. Renee, anything you want to add?

Renee Smith: Good. Thank you…

Leonard Murphy: Thank you both. this has been a great conversation. I hope listeners that you've enjoyed it as well. I guess last can people find you? I assume LinkedIn versus being flooded with emails, but All right. LinkedIn. So, just look for Frederick Charles Petit and Renee Smith that's it for this edition of the CEO series. Thank you guys so much. We'll be back again with another. All right.

Frederic-Charles Petit: Thank you so much.

Renee Smith: Lenny. Thanks,

Frederic-Charles Petit: Thanks, Lenny.

Leonard Murphy: Thank you.

Frederic-Charles Petit: Thanks. Bye.

Leonard Murphy: Bye bye.