Market Research Data Quality Crisis: Inside Rep Data’s Battle Against Fraud

Data quality is no longer a back-office concern. It’s an industry-wide reckoning.

In this episode of Greenbook’s CEO Series, Leonard Murphy sits down with Patrick Stokes, Founder and CEO of Rep Data, to explore how fraud has evolved alongside programmatic sample, why bad actors scale faster than any single solution, and what it really takes to protect research integrity today.

Drawing on nearly 20 years in the sample ecosystem and billions of survey entrances monitored through Rep Data’s Research Defender platform, Stokes makes the case that data quality is a supply-chain problem, not a vendor checkbox. The conversation spans AI-enabled fraud, the limits of prevention tech, the need for researcher feedback loops, and the leadership challenge of scaling a service-driven company without losing its roots.

A candid, ground-level look at one of the most urgent issues facing market research right now.

Transcript

Leonard Murphy: Hello everybody. It's Lenny Murphy with another edition of the CEO series. Thank you for taking time out of your day to spend it with myself and my guest. And today a man that we've been trying to make this happen multiple times and whatever the universe didn't align, but so far today it has. so I'm thrilled to have Patrick Stokes, founder and CEO Rep Data with me. Welcome Patrick.

Patrick Stokes: Thanks, Lenny. Appreciate Happy to join and be here with you. Sure.

Leonard Murphy: And thanks for your patience and all of our misfires. but here we are.

Patrick Stokes: Yeah.

Leonard Murphy: Patrick, for those who don't know why don't you kind of give your origin story, your intro

Patrick Stokes: Thanks, I'm Patrick Stokes, founder and CEO of Rep Data. I've got about coming up on 20 years mostly from the sample provider side of the industry. I was an early employee at Rewards. worked for Kurt Napton for many years, stayed there for about nine years. I'm actually born and raised from New Orleans, Louisiana. So I was working remotely for research now in New Orleans. And then in my backyard, Patrick Comr founded Federated Sample, which turned into Lucid. And I was an early employee leading their sales teams in the mid 2010s and stayed there for about six or seven years. Eventually founded Rep Data actually two weeks into the pandemic in 2020 and we just turned five years old a couple of months ago. So, it's been quite the ride growing up in two of the major industry incumbents in the early days and learned a lot from both those companies.

Leonard Murphy: Yeah, absolutely. Too. And I consider both Kurt and Comr friends. So, you've been in good company.

Patrick Stokes: Absolutely. Right leaders.

Leonard Murphy: Yeah, absolutely. let's talk about rep data. because I think the company has evolved over time. most people I think myself included, so maybe I should say myself include, but the first thing I think of repated is data quality. but I think you do more than just data quality. So, tell us a little bit about where Repata started and where it is today and we'll go from there.

Patrick Stokes: Yeah, absolutely. So, rep data obviously we want to be synonymous with providing best-in-class data quality through our but the other main pillar of rep data is service quality. So, we really hang our hat on really strong account management.  We've got a lot of tenured people in the space who provide great service, project management, quality. really where we cut our teeth is providing sample to really hard-to-reach audiences or really complex in nature projects like brand tracking or click balancing segmentations. I saw an opportunity to found rep data because I think there's just at least at the time and still today this rings true there's limited overall knowledge as companies scale and get to a certain point I think they leave behind some of the key tenants of strong account management relationship building how to be an extension of research teams we  do that through our services team. And then we've got research defenders fraud prevention product is kind of the tip of the spear.

Leonard Murphy: So, that leads us into kind of one of the topics that we never like talking about. if you guys can hear this, I have a call coming in and I'm trying to decline it. That's what happens when you go live without a net. apologies for that folks. the data quality, And this big amorphous, yes, it's like the blind man and the elephant. So, everybody's touches a different piece. They define in a different way. and no one seems to totally agree, but as you said, you're kind of at the tip of the spear. So, you're getting a pretty broad sense of no that that's an elephant.

Patrick Stokes: Right.

Leonard Murphy: It's not a tree trunk or a snake or so tell us about that because yeah I love your reports. I value the data that you're putting out because I recognize that there is so much flowing through your pipes that you are one of the companies that has one of the best views overall. Right. you're the conduit that you're the ocean that all the rivers are flowing into or many of them are flowing into. So, I want to hear your take on what that is overall and then we can talk about what we do about it.

Patrick Stokes: Yeah. What Lenny's talking about is Research Defender is a fraud prevention platform that monitors macro signals in a lot of different exchanges and panels and kind of can grade a respondent quality or not before a respondent makes it to a live survey. So, we've monitored four billion entrance over the last 12 months. And really, like to take a step back, Lenny, the early days of the panel industry, you knew exactly who was behind the screen. It was a double opt-in process.

Leonard Murphy: Right. Right. No river, no programmatic. It was Yeah. Yep.

Patrick Stokes: Yep. No programmatic. Mostly double in loyalty panels. you knew exactly who's taking the survey and that had its quality benefits but over time just from a scale cost effectiveness standpoint speed standpoint programmatic took over the space and even in the early days of programmatic there was not evolved tech enabled fraud I think over time as especially AI comes into the space and there's ic enabled bad actors. it's really critical to have a fraud prevention platform sitting on top of all panels or…

Leonard Murphy: Mhm.

Patrick Stokes: All custom market research surveys.  It would really be kind of malpractice to not have any type of protocol pre-entrant just because of the evolution of how much fraud is actually out there because right now it's a lot of open-sourced rewards apps where anyone can go, infiltrate a supplier and misrepresent themselves to gamify a survey and make an incentive,  And so it's a pretty terrifying evolution of our space and critical to really rav data raise their hand and try to be the owner of a data quality problem. but we need support from all of our suppliers. We need support from our research customers to really help us close the loop and…and own this problem.

Leonard Murphy: Yeah, I yeah, totally get that. I remember going to one of the first data quality events that Bob Letterer brought together in Chicago and hanging out with Kurt in the early days of Ewards. I mean, this is 2006, or something of that nature. so this isn't a new problem at all. To your point, it's gotten more sophisticated. and for our listeners who don't think this is some type of dirty thing just applies to research, no. When the advent of online advertising began, right, programmatic was now growth of clicks, so of clicks. So this is a digital commerce problem across the board. If there's a way to make money, bad guys are going to find a way to make money. They have forever. it's a scam. and it's a lot of money. and the problem and sorry I'll get off my soap box for a second. Let you chime in because I want to see whether you agree. I think people get bogged down. why are you doing this for 10 cents? Whatever you but just like the scale and volume of digital advertising is so significant that pennies add up to millions of dollars very quickly. so this is a cottage industry that generates the fraud component in all of its dimensions that generates millions of dollars

Patrick Stokes: Especially as the audience gets tougher to reach,…

Leonard Murphy: If not potentially billions. I've seen numbers. Right.

Patrick Stokes: Survey is longer or the incidence is lower, there's a market specific survey, whatever it may be.  If there's some type of narrow or complexity to the survey itself, that's when fraud rises.

Leonard Murphy: And the larger the reward it does as So I think there's significant challenges B2B. So all right. All right. So, we have this big problem. and I talked about this a little bit beforehand that this and maybe it helps everyone think about this. This is a supply chain issue, right? So, if we were thinking about any other product in the world, there's, whatever the core product is, you're cutting down the wood, Is it good wood? then you have to mill it. And does that happen, Is the board straight? Is whatever the case may be, right? We can think about it. there's multiple phases and constituents involved in the supply chain till you get to the final product of that and anything else in garbage out. So your point is we don't look at this necessarily as an ecosystem from a supply chain standpoint that all has a role and responsibility in ensuring quality. Instead it just kind of gets kicked down the road, Nobody wants to take responsibility for this and they see it as someone else's problem.

Patrick Stokes: Right? Yeah. I think at least the comparison you made is a research survey it could be written too long or the can have too many open ends there's a lot of variables really fraud is a cat and mouse game I mean research defender is blocking almost 30% of all entrance at some point you re reach kind of like a critical question mark there of If you block even more,…

Leonard Murphy: Why not a scary question?

Patrick Stokes: Can you even fill these surveys, So what we really need is like a buy in from not only our suppliers but our research partners to help us close the loop because again it's a cat and mouse game. Rev Data and our product research offender were raising our hands saying we're going to try to own this problem as best we can.  The reality is tech enabled fraud evolves faster than any fraud prevention tool out there. So, we need to rely on our research partners to help us close the loop on what they're seeing in the survey data that we can help take a look at and tie back to the respondent, tie back to the pre-entrance signal we have so that we can continue to build better quality assurance layers in the future. So, I think it's a co-owned solution between the suppliers, Rep Data, our market research partners to really help tackle this. It's bigger than, just we can't just say the supply partners only own this. It's like a market research overall problem that needs buyin from all different types of stakeholders.

Leonard Murphy: Yeah and as objectively I get all that and I look at everybody around and there's a bit of a pissing match going on, there's a one group that wants to own this there's another you have competitors from a technology standpoint and that's just competition great fine got it however what I hear you talking about is that effectively we need a central repository and clearing house of data that we could all learn from and then fine pick your solution right which one that is. And that just seems like an uphill battle to get anybody to adopt any level of integration to share information. I tried it with Baragliff. So I'm sure you remember back we were going to connect it all and be the central supply chain for data through our pipes and there be a quality component and nobody was going to do that. so I feel your pain. My point is so when not only is we have a fragmented ecosystem with competing solutions it makes it hard for there for anyone or hell everybody to get a handle on this rapidly evolving space. It seems like it's a game of whack-ole that nobody could win. Because there's just too many people trying to do it all at once and Is that a fair assessment? I'll give you that. Yep. Yep.

Patrick Stokes: Yeah, I think the scale of 4 billion scans helps us have the best viewpoint into the market trends. I'm really asking out there for market researchers is to arm us with what they're seeing and maybe questionable data in their surveys itself so that we can do some type of feedback loop look back to spot emerging fraud trends right and…

Leonard Murphy: Yeah. Yeah.

Patrick Stokes: So I mean I'll give you an example we were doing some market specific work for a customer they noticed a spike in a really unique type of fraud we'll call a batch entrant surge of respondents. And what we looked back at was and we found was a new type of fraud where the bad actors were using a unique combination nefarious activities that we had not seen before and…

Leonard Murphy: Yeah, sure. Yeah.

Patrick Stokes: We had to update our technology to catch the specific markers that we had not seen before. But we don't get that ability to evolve if we don't hear from our customers and get data to tie back to the pre- entrance, right? And so that's Another example is monitoring specific devices or operating system trends and chat GBT open AAI operates on a very specific system right and we can kind of see through the data the emergence of very specific AI based operating systems right and…

Leonard Murphy: Yep. Yeah.

Patrick Stokes: Those are just a couple of examples but we need the feedback loop from the researchers Because what we have is the pre-entranting.

Leonard Murphy: Right. Yeah.

Patrick Stokes: We can see all the scans and we've got a lot of different signals and thresholds for appetite of what qualifies as a pass or a fail. But ultimately, again, it's a cat-and- mouse game that these bad actors are going to try everything they possibly can to circumvent any fraud prevention, any survey to get that incentive at the scale that you're talking about, Lenny. and so for us, partnering with our researchers to help us get smarter is critical as we grow our presence in the space.

Leonard Murphy: Yeah. God bless, I mean, you try to tackle a big old issue, that's complicated and one that nobody wants to deal with fundamentally,…

Patrick Stokes: Yeah. Right.

Leonard Murphy: I mean this isn't it it has all types of implications of but it can't be ignored u to your point and it is only getting worse it does bring up an interesting question now we'll kind of transition to something else though so I think about this all the time right we are rapidly emerging into if not already there in a world where a digital clone of an individual very well could be appropriate to take a survey. So let's say in the world of agents, right, if I have a shopping agent, which I don't, but there was a Lenny shopping agent, right, that's trained on all of my buying behaviors and preferences, etc., you can make a strong argument that having that AI take a survey on shopping behavior is 100% accurate and appropriate even though it's not me as a person. so I bring that up just because we're in this weird world, But you're looking at those type of things. a point of view on is it just going to add insult to injury, a whole other level of complexity that we have to go

Patrick Stokes: Yeah, I think the complexity is really around is it really Lenny that they're building the twin on or…

Leonard Murphy: A right. Yes.

Patrick Stokes: Is it a fraudulent respondent that circumvented a panel and they're building their models based off of fraudulent data. And so that's the big question there. But certainly I think the industry embraces any type of evolution looking from double in panels to programmatic to agile insights to lot of resc platforms out there using AI to its advantage. So we want to embrace the market and the evolution and I think Rep Data is uniquely positioned with our fraud prevention suite to kind of move as the market moves.

Leonard Murphy: Yeah. Interesting times. Interesting times. All right. So, you been in the industry for going on 20 years. you launched a business during the pandemic, which just the best time in the world to launch a business, what have you learned and you decide to tackle a big old nasty hairy problem that's not easily defined. So you set some pretty lofty goals for yourself. what have you learned as a CEO

Patrick Stokes: Yeah, I think I learned a lot from Kurt and Patrick Comr. I mean, Research Now went from $30 million a year in revenue to $300 million a year in revenue. And it's very challenging to maintain a culture…

Leonard Murphy: Mhm.

Patrick Stokes: A startup culture as you scale.  It's also very challenging to evolve and disrupt with technology like Cint did or Lucid did. And so I kind of learned I think the best of both worlds, how to operate a fast growth services business from the early rewards. I learned a lot from Kurt there and how to kind of build and grow a technology- based company like Lucid. So, I think rep data has the best of both from I think what I've really learned is that I love growth, right? I hear you.

Leonard Murphy: Patrick.

Patrick Stokes: Yeah, Can you hear me, Lenny? Yeah.

Leonard Murphy: Yeah, you broke up just for a second, but that's okay. Yeah, I heard I love growth and then your image froze and you fade out for a second.

Patrick Stokes: Yeah, maybe we can end the call on that. I love growth. I'm a growth minded CEO. something's going on with the computer here, but I wonder…

Leonard Murphy: I can't hear you.

Patrick Stokes: If it's the invite for the fourth time, but I think I don't know if you can hear me still.

Leonard Murphy: And…

Patrick Stokes: Yeah, sorry about that.

Leonard Murphy: Maybe that's okay.

Patrick Stokes: Yeah, I think anyone who's been around me really appreciates my growthminded mindset and…

Leonard Murphy: Go ahead.

Patrick Stokes: My passion to win. I'm very competitive. really like to provide world-class service in the world we live in, Lenny. there's a lot of competitors,…

Leonard Murphy: Yep.

Patrick Stokes: There's a lot of choices. it's really about do you like the person you're working with? Do you trust what they stand for? can you work with this team? Are they in it with me? And I think some of the lessons with that I've seen over 20 years is we can't forget about the little things. We can't forget about how we got here or why we're here and what we stand for. What are the original roots of the company? And so I want to just maintain the mission of Rep Data to be reliable and repeatable for life. That needs to hold true as we scale, as we grow. And so that's the real challenge of being a CEO is it's easy in the early days where you're sitting next to your employees and you have connectivity to every single account, every single project going on, every project fire, who's happy. that's really easy. It's a one or two year old business. but as a 5-year-old scaling business, probably one of the fastest growing companies in the space, you really have to start to rely on your leadership team and trust the process, trust what you've built and m try to maintain connectivity to the original roots of the company as best as possible.

Leonard Murphy: Yeah. Yeah. I think that is definitely a powerful lesson learned. All right, we'll wrap up on kind of the final question. What's ahead? What are you excited about as we look go I can't believe we're almost at the end of 2025. So let's say for the next year. what do you think? All right. That's going to be cool.

Patrick Stokes: Yeah, I think for us with Rep Data, we're at an inflection point. We partnered with Mountain Gate Capital in March. We've got a great aligned strategic plan for growth. we're going to evolve our fraud prevention capabilities, continue to double down on service quality, maybe build, easier bridges for enterprise brands to connect to rep data, but continue to work with our agency partners to help close the loop on evolving exciting, stressful times to your point. It's a big problem that we're trying to solve, but we're here for it. we are raising our hands to be the data quality thought leader in the space and that comes with a lot of high pressure but potentially great outcomes.

Leonard Murphy: Absolutely. Fool's leap or angels fear to tread could be one of the outcomes here or you're the hero. so far it looks like you're going to be the hero. So I look forward to staying ab breast of…

Patrick Stokes: You got it.

Leonard Murphy: What you're doing again. And I have to say, love your transparency how you've been taking this data and putting it out there. If you guys are not following Patrick or his team or just rep data on, LinkedIn or other social media, you really should they are sign up for their newsletter, getting their marketing list because they are routinely sharing really interesting and important information about the volume of data flowing through their pipes. Or better yet, here I'll give you the plug, Patrick. Sign up, be a customer, right? …

Patrick Stokes: That would be great.

Leonard Murphy: But yeah, but if you're not in that role to make that decision, do follow along because it is really valuable and I've definitely appreciated the information you've put out there. All right, Patrick, thank you so much for your time. glad that we pulled this off. Where can people find you?

Patrick Stokes: Yeah, I'm on LinkedIn. Also, if you want to send me a note, Pete Stokes at Rep Data. I'm always happy to have conversations and, I think we've got a really good team here to help any way we can.

Leonard Murphy: All right.

Patrick Stokes: Right.

Leonard Murphy: We will end on that note. Thank you to our listeners. Thank you to our producers and sponsors. we really appreciate all the time. That's it for this edition of the CEO series. Everybody take care.

Patrick Stokes: Thanks. How you doing?  Yeah.

Leonard Murphy: Bye bye.