Conveo’s Hendrik Van Hove on Turning AI Into a Research Superpower

Conveo co-founder & CPO Hendrik Van Hove joins Lenny Murphy to unpack how AI is reshaping market research—from the rapid normalization of AI moderation to new agency business models and always-on consumer access. Hendrik traces Conveo’s origin (McKinsey to Y Combinator), the importance of bringing veteran researchers into the loop, and why enterprise teams are shifting from “fewer, bigger” studies to many rapid, compounding projects that speed impact (e.g., Unilever sprint examples).

The conversation ranges into what’s next: privacy-centric data ownership, wearables that enable seamless voice/video qual at the shelf, and building living consumer models (digital twins) that connect research directly to decisions. If you’re navigating AI adoption, panel strategy, or the future of insights workflows, this one’s a blueprint.

Key Discussion Points:

• From idea to YC: Why research was a “perfect AI use case” and how Conveo blended tech + MRX expertise.
• AI moderation is table stakes: Market skepticism flipped to acceptance in under a year—and what “better” really means.
• Agency & enterprise transformation: More projects, faster iteration, and business-model shifts beyond cost/speed.
• Data ownership & privacy: Toward participant-controlled data and monetization as passive/voice inputs expand.
• What’s next: Wearables, always-on access, and “digital twin” consumer models that compound learning over time.

Resources & Links:

• Conveo
• Y Combinator
• ESOMAR
• MRII
• GRIT Insights Practice Report
• Meta Ray-Ban smart glasses

You can reach out to Hendrik Van Hove on LinkedIn.

Many thanks to Hendrik Van Hove for being our guest. Thanks also to our production team and our editor at Big Bad Audio.

Transcript

[00:00:10] Lenny: Hello everybody. It’s Lenny Murphy with another edition of the Greenbook Podcast. Thank you for taking time out of your busy day to spend it with myself and my guest. And today I am joined by a guy I’ve had a chance to get to know over the past couple of years. Full disclosure, had a brief commercial relationship as they were exploring things. But Hendrik Van Hove, the co-founder and CPO of Conveo.ai. Hendrik, welcome.

[00:00:37] Hendrik: Hey, super excited to be here and great to see you again, Lenny.

[00:00:40] Lenny: It’s good to see you as well. It had been too long. So, I’ll build off that background a little bit because I think it’s relevant to the context here. So, we all remember the great AI boom, right, when suddenly we started seeing commercial applications of generative AI being [unintelligible 00:01:00] research space, especially within qualitative research. And one of the early companies in doing that was Conveo, and you had reached out to myself and Greg Archibald at Gen2, and said, “Hey, take a look and tell us what you think.” Which was very cool and very exciting to be able to do that, and also incredibly instructive for me to realize, holy crap, the game has just changed, you know? That the applications of these platforms, even it’s early iteration, right, when you were just kind of coming out of beta. You know, maybe not… certainly not as sophisticated as it is now, but it was that clearly the writing was on the wall, on what we were going to see. And also amazing, you and your team coming in at that point with no research background. So, I want to make sure we just kind of dig into that. So, let’s use as a segue. Let’s talk about, kind of, your origin story, your background, and how you got to Conveo, and then we can dig in and get a little more geeky on other stuff. 

[00:02:15] Hendrik: Yeah, awesome. So, as you mentioned, I’m Hendrick, co-founder, and I do basically everything commercial, so try to spend as much time with our customers as possible. And I’m not technical, so I can’t really, let’s say, build a platform, but I did study artificial intelligence at university, and so from there on out, I basically started working at McKinsey, and right before that, ChatGPT launched publicly. And yeah, it was all a frenzy, right? And so, I was super lucky because I started at McKinsey, and there they were like, “Okay, you studied AI. You’re now going to do all these AI use case scans for Fortune 500 companies.” And basically what that meant is sitting together on a board level and the C-suite, but especially sitting next to practitioners on every side of the business. So, that’s marketing, research, sales. And one of those use cases that really stood out to me was research. I also did a ton of B2B interviews at McKinsey, but that was mostly very expensive experts, so not the same as what we do today. But to me, market research was just a perfect use case for AI, two years ago already, because there were a lot of shiny things that really didn’t work at scale. But it’s also a problem that can never be solved, so you can keep building and creating value for years to come because you can never understand your market perfectly. And so, that’s really how it started, and that’s now two years ago, and here we are.

[00:03:51] Lenny: That’s fascinating to be in that position of, like, oh, I’m being paid to figure out a new business. Um, okay. The—[laugh] from McKinsey. And that’s often, even in my role today, that there’s similarities of, I’m looking at all of these things you know, that are going on and talking to folks. And there’s always a back of my mind of, oh, there’s an opportunity there. You know, no one has identified that gap. You took it a step further of you went into Y Combinator. So, how did you get from, “Hmm, maybe there’s something here,” to actually getting selected in Y Combinator and going, you know, starting to bring this to life.

[00:04:38] Hendrik: Yeah. So, to us, Y Combinator as a Belgian company, right, it’s really a game changer. Like, as soon as you get in, you get so much more access to venture capital, et cetera. It’s really, like, also a confidence boost to, let’s say, like, okay, hey, we are here and we can really go for it. But so how we got in actually applied in, I think, October 2023, and then we didn’t get in the first time. So, to me, I had enough confidence in the idea, Y Combinator had also said, like, hey, keep working on the idea. So, quit McKinsey. Sat around the kitchen table for a few months. Basically met my co-founder at Dieter, who’s hyper technical. He built a $500 million company before as CTO. So, we then met and started building the platform. Six months later, we applied again and then we got in. But as you mentioned in your first question, yeah, Dieter was a CTO. I have a business background, basically some understanding of AI, but we didn’t really have the market research knowledge, right? And so, we got in, but then we quickly realized, say, if we’re going to want to do this, we want to do it well, so—and then luck had it, actually, that we met Niels at a wedding, via an introduction and one thing turned to another, Niels immediately said, after two, three meetings, say, “I’m going with you guys to San Francisco.” And so, for those who don’t know Niels, maybe you’re better positioned to give the introduction, but he’s our head of market research. He’s been in industry for more than 25 years. Founded Human8, and was president at ESOMAR, involved at MRII, et cetera. And with the three of us, we really have what we need to make the AI agents be really like a research assistant. Niels is able to instill all these frameworks, et cetera. And if you hadn’t had Niels, then it’s so easy to hack together an AI moderator, but to make it really good, with these time-tested techniques that the market research industry is built on, that’s actually extremely hard. And so, yeah, you’re very right to ask, like, I’m an outsider, but there is incredible value in having a market researcher on the team who builds the AI into what it should be.

[00:06:49] Lenny: Well, let’s explore that a little bit more because I think, you know, in the initial explosion of tools like Conveo and your competitors, there was a real threat feeling from a lot of folks in the industry that you know this is going to take away my job. And in some cases, maybe so, but in most cases, I think it quickly became apparent that, no, this is a superpower. This is an augmentation and when built correctly, fit for purpose, as I would say that you guys have done with Neil’s guidance, that it creates new possibilities that we previously did not have, particularly at scale, right? I mean the speed, cost, efficiencies. I mean, those are no-brainers, which obviously you saw early on as part of your idea. So, now you’re in the market. You received a substantial VC round as well, so you know, a lot of confidence from folks to say, yeah, we think this is great, and this is going to be transformative. How has that—are you seeing in the market, that shift from fear of, like, holy crap, no, this is going to kill me, and you know, eliminate my job, to, oh, this is a superpower, and it’s going to enable me to do things differently. How is that happening right now, kind of, in the trenches?

[00:08:20] Hendrik: Yeah, there’s kind of a change we went through internally. And then there’s the change you’re describing. The way that I really felt it was a year ago, and it was only a year ago, and our initial product was getting ready to sell, there was so much skepticism still about AI moderation. Like, we had to convince people there was no follow up, et cetera, and yeah, it was just we needed to really find those early adopters and early believers, but most people still dismiss the moderator. Today, no one, even when we do a demo, wants to look at the moderator. They’re just like, yeah, AI moderation works. And in less than a year, that shift has been absolutely incredible.

[00:09:02] Lenny: That’s crazy.

[00:09:04] Hendrik: Yeah, really, it keeps blowing me away. So, that’s a very big one. Then the second thing we went through internally is, we’ve kind of started in the market with a solution, right? We said, like, hey, we have this, an AI moderator. You can now do quality at scale, et cetera, but we spent a ginormous amount of time with our researchers and their stakeholders, more and more, and we’ve kind of fallen in love with the problem of having insights and understanding your target markets and drive revenue for the business. And so, whereas the AI moderator was very much our initial focus, we are now much more focused on, hey, how do we solve this problem? And that means also being able to, for example, upload videos, we’re experimenting with building social listening, learning across different studies, et cetera. And so, that comes into your question like, hey, are people actually scared of losing their job, but if you give people all of these powerful tools to generate revenue impact in the business, which you get because it’s so much easier to run research that you’ll just run more of it, right? You won’t replace researchers because everyone has a backlog of the 101 projects, then actually, your role becomes way more relevant. And so, researchers at Unilever, for example, they have a three month sprint. Instead of running three projects, they now run 10, 15, 20 projects with Conveo, and their stakeholders are loving it because all of a sudden, instead of a blocker, you’re an enabler. So, I think researchers have yet to see their best days.

[00:10:47] Lenny: Boy, a lot of follow ups on that.

[00:10:48] Hendrik: Yeah, yeah go [laugh].

[00:10:49] Lenny: Well, is one, what you said this year, that’s kind of what we detected with GRIT. You know, we use that as a tool. It was about this time last year that it shifted. So, it was just very clear in the data, that we went from, you know, early adopter skepticism to okay, you know, we’re going to embrace this. And it’s still maybe somewhat marginal, but—not marginal, but there’s still ways to go. So, that shift is tangible. Multiple data points—

 [00:11:19] Hendrik: Exactly.

[00:11:20] Lenny: —have validated that. And that idea that tools like this transform the value that we can deliver, not just speed and cost efficiencies, but how does that, you know, the cheaper, faster, better, how do we zoom in on what is that ‘better?’ And from what you said, it sounds like that your clients are recognizing that better component is not just speed to insight, but speed to impact. For a company like Unilever, you know, okay, we’re we get to be very iterative, get to be very agile and build better products with a greater likelihood of success. And so, you’re looking at that systemic transformation that is proving itself as this is tangible in real-world business impact, moving the needle for your users in creating better products—

[00:12:26] Hendrik: Exactly.

[00:12:27] Lenny: —or messaging. Yeah, you know—so we’re recording this on October 9th, and last week was SMR. And were you at the SMR conference?

[00:12:40] Hendrik: Yes. We had a great showing, talk with Unilever, even. It was, yeah, a great conference.

[00:12:44] Lenny: Okay. So, I heard through the grapevine that there was a panel with Simon Chadwick and Kristin Luck, and they name-checked me, and Simon said, “Well, Lenny thinks the industry is doomed.” Now, come to find out that is not the full context. The context was, he was talking about conversations he and I have had around this very topic. The transformation of the process has implications for business, cost, you know, all of those things. And so, companies who don’t adapt very well may have real challenges, right? Business models are shifting as well. So, you’re using an end-client example for Unilever, but I assume it’s the same thing goes to your supplier-side agency users that, look, this is a superpower that it’s going to change, but you can now do more than you could have done before, right? It’s an augmentation. It’s, you know, a multiplier effect. But the implications are the way you’ve been organized and even the way you were charged, your revenue models, now must shift. Are you seeing that as well? Are you seeing that transformation on the agency side, users are looking at utilizing Conveo as competitive advantage as they transform their models as well?

[00:14:14] Hendrik: Yeah, definitely. I think agencies always feel it first, right? You’re always in a bake-off against the next agency that has a shiny thing, and there’s so few barriers to changing or trying something new for clients, so agencies always feel it first. And we really try to listen super closely to our agencies and say like, “Hey, how can we help you win more pitches”—right, because that’s where it all starts—“And then deliver the actual value to the client.” And just the other day, I got an email back from an agency, and I absolutely loved it. The title was, “Aha, now I get it.” And so, what the person wrote was, where we use Conveo is not to replace the main data collection point that we would traditionally run as an agency, but we now have, like, instead of one big data collection sprint, we maybe have 5, 10, or even 20 small data collection sprints around it that we do with Conveo, sometimes even only five people, right, just to get a feel. Or you do, like, post-interview or pre, et cetera, and they can just have many more valuable insights and touch points with their clients. And so, it’s really about not only about saving costs, and yeah, a client will cut your costs down. No, the client has their budget, right, and sure, sometimes they cut it, but ultimately, it’s a competitive market and they need to have better insights in their competition to defend market share and generate more revenue, so the more value you can provide as an agency for the same time, yeah, the more you’ll be in vogue. And so, that’s really the agencies we see leads. They are super creative in terms of what they do for their clients with these tools.

 [00:15:55] Lenny: Yeah. Yeah, I think that we’re headed into kind of the golden age of the industry. And for the record, Simon and Kristen and anybody who was in that SMR panel, I agree. I just think it’s also incredibly evolutionary, driven by the type of imperatives that you’re talking about. All right, so let’s get a little more into the weeds. So, you have, [laugh] I won’t ask you about a coding or anything of that nature, but so you had this idea, you brought in experts to help refine it, and get it fit for purpose, you’re getting market acceptance. Now, there’s also these bigger issues that we’re grappling with, the whole world’s grappling with, around data ownership and privacy, particularly as you’re evolving models, also on building data ecosystems that unlock even more value from the data, you know, through connectivity and other systems, directly into LLMs. How are you guys navigating that? Have you found those to be challenges? Is this still something that you’re working through? And what do you think the future looks like, when we think about data privacy and ownership, while also data being the fuel that drives more innovation?

[00:17:21] Hendrik: Yeah, two thoughts there for me. I think this is one of those things where people can overestimate the one-year progress and underestimate the ten-year progress. And so, to us that we’re actually not even thinking about it right now. We are doing the AI moderated interviews, people have videos that they want to upload, they’re asking us to do quants, we are doing a lot with the data processing and bringing that closer to concept generation, and we have, like, the multimodal analysis, et cetera. Those are all things that are incredibly new, but in terms of connecting with more continuous data streams and new data ecosystems, et cetera, that’s not something that’s at least on our client’s mind. So otherwise, we would prioritize it because that’s really how we work. Now, if we zoom out and if I see that two weeks ago, Meta launched their new Ray-Ban glasses with a screen in it, et cetera—I’m maybe going to take you on a little trip here—but if everyone, if OpenAI, and all these AI companies want to give you a personal assistant, right, which is obviously their goal, and I would love one, then that personal assistant is going to be as valuable as possible if it knows what you see, what you are doing, what’s even on your screen, most likely, and so it’s be always on, gathering context. What I think is going to happen is, instead of asking what a panel will look like in the future, will be the an app, the Dynata app or something, or maybe it’s Conveo, I don’t know, and it will say like, “Hey, do you want to participate in this panel for I don’t know, Reckitt, P&G or whatever?” We’ll look for moments that are interesting for us, and once you are in such a moment, we will give you a little notification and ask you a few questions on how you make your decisions. And most likely, you will probably already have shared why you made those decisions, right? And so, then you are looking at a completely new way of tapping into many more people because all of a sudden, everyone is addressable. There’s much lower effort from the participants to share because most of the info the AI already knows, so you’ll just say, hey, sure, I’ll share level three information data of this week, with P&G at a much lower cost, and for the brands, it’s much more relevant and timely, and it’s not claimed behavior, it’s observed behavior. And to me, like that’s going to be… that’s going to take shorter than people think before we’re in a world like that. And the impact— 

[00:19:57] Lenny: That’s already happening.

[00:19:58] Hendrik: Yeah, it’s already happening. Exactly.

[00:20:02] Lenny: Yeah, yeah. Sorry, I didn’t talk over you, but I agree with you one hundred percent. For those who aren’t watching the video, I was nodding vigorously because those conversations, in my consultant role, I’m privileged to talk to lots of different folks, and I am keenly aware that money is being spent right now to bring that overall vision to life, and in many cases, it already is alive. It is happening. The old William Gibson quote, right, ‘The future is already here. It’s just not evenly distributed yet.’ So, yeah, I think that I agree with you. But you started with talking about Meta and the Ray-Bans and wearables. I was an early adopter back in the day, one of the first people to try that technology. I’ve always believed in the potential of wearables, and now that infrastructure issues like bandwidth and processing power are behind us, the focus is shifting again. Are you thinking about how engagement will look when it’s voice-based or projected through wearables instead of screens?

[00:22:15] Hendrik: We’ve chosen to focus on voice and video only — no typed answers — because the quality is much higher and it aligns with where the future is heading. The Meta Ray-Bans, for instance, allow subtle, hands-free interaction. Imagine you’re in a supermarket, wearing smart glasses, and an AI prompts you: ‘Can I start a quick interview with you about this shelf?’ You respond naturally and continue your day seamlessly. That’s the kind of real-time, frictionless engagement we envision.

[00:23:36] Lenny: For a company like Conveo, the move from structured surveys to conversational formats is huge. Traditional quant surveys made data easy to analyze but limited depth. Conversational formats have always faced scalability challenges, but AI is removing those barriers.

[00:24:10] Hendrik: Exactly. And in the future, we may not even need people to actively answer — AI could gather data passively and give users control over how that information is shared or monetized. That’s a better, more transparent model than the current system where users give away data unknowingly.

[00:25:42] Lenny: I completely agree. Back in 2017, I tried to build something similar — a system where personal data is treated as an asset. The idea didn’t take off then, but I still believe that’s the direction we need to move in. AI can finally make data ownership simple by managing decisions for users and making personal data a natural digital exhaust rather than a chore to control.

[00:27:33] Hendrik: Exactly. The challenge has always been the balance between effort, privacy, and benefit. Voice and video data are richer and more valuable, especially as AI begins making personalized decisions. When that value increases enough, people will finally want ownership of their data.

[00:28:39] Lenny: You’ve made me realize I was wrong when I said Conveo couldn’t become a unicorn. With this vision — building the enabling infrastructure for real-time conversational data and ownership — you absolutely can.

[00:30:09] Hendrik: Right now, Conveo is focused on evolving research workflows — blending quant and qual, automating moderation, and improving cost and speed. But yes, in the long term, solving the global panel challenge through AI-driven recruitment and engagement could unlock that larger vision.

[00:31:22] Lenny: Exactly. There’s already movement in that direction — major AI companies are looking for real-time consumer data feeds to power their models. It’s only a matter of time before they make plays in that space.

[00:33:37] Hendrik: Totally. LLMs started as text-based, but now we’re entering the multimodal era — voice, video, and real-world context. The next frontier is real-world AI: robotics and personal assistants that understand lived human experience through contextual video and audio data. That’s where this is all heading.

[00:35:23] Lenny: Absolutely. When Meta bought Instagram, it wasn’t just about photos — it was about training visual recognition models. The same logic applies now with wearable data and multimodal AI.

[00:35:41] Hendrik: Exactly. Today, every major tech acquisition and product decision is driven by data strategy. Even apps that seem trivial are built to train models or collect multimodal data. That’s the real play.

[00:36:13] Lenny: It’s all about unlocking the value of data and creating systems that can make use of it efficiently. Platforms like Conveo can enable purpose-built, focused data ecosystems for brands that want defensible, proprietary data assets — not just aggregated social data.

[00:37:45] Hendrik: That’s what we’re obsessed with. Today, insights lack longevity. We want to help brands build living, evolving consumer models — effectively, digital twins of their markets — by connecting every interview and dataset. This bridges the gap between data lakes and decision-making.

[00:39:26] Lenny: Completely agree. As we centralize and integrate data, the number of questions we need to ask will decrease — but the importance of those remaining questions will skyrocket.

[00:39:44] Hendrik: Exactly. Companies will move from one-off studies to continuous learning ecosystems. Each project builds on the next, feeding into a living, voice-and-video-based knowledge model that evolves with the market.

[00:41:18] Lenny: Yes, and the pace of that evolution will be much faster than any previous transformation in insights.

[00:41:30] Hendrik: Exactly. Which is why it surprises me when large organizations try to build everything in-house. Speed, scalability, and adaptability come from collaboration and focus, not isolation.

[00:41:58] Lenny: Right. We could go on for hours, but let’s move toward wrapping up. Is there anything you wanted to cover that we haven’t touched on?

[00:42:24] Hendrik: I could give a pitch for Conveo, but really, the biggest lesson I’ve learned is that success comes from building faster — and smarter — than anyone else. Our engineers talk directly to clients, hear their challenges firsthand, and build immediate solutions. That’s how innovation happens.

[00:43:53] Lenny: That’s a great philosophy. I tend to think big picture first, but you’re right — the real progress comes step by step, focusing on today’s problems.

[00:44:16] Hendrik: Exactly. Being too early can be as risky as being too late. I’m a futurist at heart, but success requires timing — building sustainably until the world is ready for the next leap.

[00:45:27] Lenny: Well said. Always happy to brainstorm the future with you, Hendrik. For those listening, how can they reach you or learn more about Conveo?

[00:45:53] Hendrik: You can find us at Conveo.ai or email us at [email protected]. We’d love to connect and continue the conversation.

[00:46:06] Lenny: Okay. Well, any last thoughts you want to get out there before we wrap this up? 

[00:46:12] Hendrik: No. I truly loved the conversation, so thank you for your time as well, Lenny.

[00:46:16] Lenny: Always. Well, thank you. So, I guess that is it for this edition of the Greenbook Podcast. First, big shout out to our listeners. Thank you. Obviously, Hendrick and I find sometimes to talk, but we hadn’t in a long time, so I appreciate you giving us the excuse to do this, and hopefully you found some value and inspiration out of our rambling. We do know that you have choices, and making that choice to spend time with us, we appreciate very much. Big shout out to Brigette, our producer. Without her, none of this would happen. To our auditors, Big Bad Audio. To our sponsors, which I think Conveo you’re actually going to be, are one of those sponsors, so thank you again for that as well. It’s always nice to do something that you’re actually helps pay the bills, so thank you. And with that, we’re going to sign off for this edition. We’ll be back with another Greenbook Podcast real soon. Thanks a lot. Everybody have a great day. Bye-bye.