The Synthetic Respondent Problem Nobody's Talking About

Check out the full episode below! Enjoy The Exchange? Don't forget to tune in live Friday at 12 pm EST on the Greenbook LinkedIn and Youtube Channel!

The research industry is having a reckoning with trust. Fresh funding is pouring into measurement and event intelligence, Nielsen just walked away from MRC accreditation, and AI-moderated qual is forcing a hard question: when is talking to a synthetic respondent good enough, and when does it quietly undermine the answer you're paying for?

Thanks to our producer, Karley Dartouzos. 

Use code EXCHANGE to get a 15% discount on your general admission IIEX tickets!

IIEX West 

IIEX.AI 

Stay Ahead of the Curve! Subscribe to The Exchange Newsletter on LinkedIn Today! 

Transcript

Karen Lynch: Oh, I'm still talking.

Lenny Murphy: That's all right. It's Friday. We all got to take a drink of something. Depending where you are, who knows what you're drinking?

Karen Lynch: Seriously, this is water. This is water. It's too big to not be water.

Lenny Murphy: This is Snapple. So the diet, the zero sugar snapple.

Karen Lynch: So beverage of choice, beverage of choice.

Lenny Murphy: Yes. All right. We have switched, by the way.

Karen Lynch: switched. I like to share this to zero sugar zero Coke Zero instead of Diet Coke partake. Coke Zero, man, right?

Lenny Murphy: It tastes like Coke.

Karen Lynch: It tastes like Coke. And it's so funny because we don't drink a lot of it, but when we choose to have a soda in this house, it is Coke Zero. All of us. It's like universal now. Anyway,

Lenny Murphy: I agree. And there's. Uh like Dr. Pepper Zero that's better than dyeing Dr. Pepper. That yeah, the zeros taste to me, they taste better.

Karen Lynch: Very interesting, anyway. Yeah, that's me digressing.

Lenny Murphy: We do, we do, we do. Um, uh, and as always, there's a lot to go through. Uh, so much, so much the warning for everybody again, so like we condensed probably easily 50-60 things, and easy, easy.

Karen Lynch: manageable.

Lenny Murphy: So, keep that in mind. We're trying to keep, we're trying to give love to as many people as we can in the stories, but there's a lot.

Karen Lynch: There's so much, too much. Yep. No, not too much. There's saying too much is not the right thing to say, but a lot is the right thing to say.

Lenny Murphy: A lot, a lot. Uh, speaking of a lot, the next Tuesday, the grip form. Um, and uh, did we agree that you are going to bless?

Karen Lynch: Us with your presence did we agree did we agree that I'll join you no matter what yes I think we agreed this is how lenny and I talked just so everybody knows this is this is what life is like I think the first two I'll join you no matter what um real-time decision making happening here um yes I can I'll join you for the first two but the last one it you're there's be five of you you don't want it to be six that it you know, it'll never end, it'll never end.

Lenny Murphy: So, good news is that usually you guys have to meet everybody who's attending the grit forum, which you should absolutely attend. Um, you have to listen to me. Um, now you get to listen to Karen and me. It's like, how awesome is that? Doesn't get any better than that, plus, our wonderful panelists. Oh my gosh, it's so fun! Yeah, and there is a lot to talk about, guys. The um, uh, we will definitely, because what if you haven't attended the grit forum, obviously, the panelists we break it up into different sections thematically, panelists talking views on specific topics and sections within grit, but then we also tie it all together, uh, often with recent, you know, okay, great. I mean, grit is only we collected those data three months ago. Um, and then a whole lot has happened in three months. So, uh, so do pay attention. Please attend. Um, it's great, it's fast-paced. Uh, lots of others, lots of very smart people. Um, and then this week, next Tuesday, we will be blessed with Karen, with Karen participating in panels. You know what happens when you postpone a week, all things you know, things change, and you never know who's going to show up. It gets better, it gets better.

Karen Lynch: People. All right.

Lenny Murphy: And they're speaking of, well, different, I guess, and better in its own way. West now, big event.

Karen Lynch: Oh my gosh. When you know how excited I am for IIEXWest, because what's happening is I am, you know, continuously curating these Speaker spots and I'm almost done. But the brands that are showing up for IIEX West at this point, because again, it's you know, the speakers that are taking to the stage, I just kind of can't get over them. They're, um, we are being joined by some amazing brands. Somebody from Visa is showing up: Airbnb, Meta, Microsoft, Chipotle, you know, Rocket, Clorox, Mozilla, Roblox. It's like the list is going on and on. And yeah, you know, ancestry, meta. Eli Lily. Come on. I mean, it's just amazing. We have some amazing brands that are going to be showing up at this event. It's going to be an epic. Absolutely.

Lenny Murphy: Epic. And different for those who think, oh, it's just another IAX. It is not. It is structured differently. So try to get the core of what makes IIEXspecial, but given its own unique breathing room and vibe.

Karen Lynch: Yeah, and like the little bit of context, you know, we're bringing back round tables at IAX, which there have been in the past. So roundtables at IIEXaren't necessarily new. What's new is that I handpicked table hosts. From my repertoire of QRCA moderators primarily. So, and some die-hard industry experts like Will Leach is facilitating a roundtable, Z Johnson and Rob Volpe, and then some folks that I've just known forever in the moderating world, you know, Katrina Noel and oh gosh, Mark Engel. And you know, I just could go on and on, Gary Redmond. And there's again. Like, there's going to be eight different topics each day. So 16 total covering anything that people might want to talk about, whether it's synthetic data or storytelling or data quality or, the convergence of UX and market research and the ROI of insights, like so many great topics, it will be hard to choose what to sit and discuss. And if you don't want to go to a round table, just sit and have a longer lunch. I mean, on the pier in San Francisco. I just think, like, anyway, if I was an attendee, there'd be no better event.

Lenny Murphy: Yeah, I'm not hearing any downsides. What's wrong?

Karen Lynch: No, there's no downside to this one. Plus, the competition. I mean, come on.

Lenny Murphy: Yep. Yep. All right. Well, speaking of competition, shall we dive in? Yeah. You know, the train just keeps on rolling.

Karen Lynch: It really does. It really does. And yeah, yeah, go ahead. Kick things off with this. I don't, yeah, I don't know ABC's insights. ABC.

Lenny Murphy: So this was Nielsen. My understanding is that there was a JV called Nielsen Catalita. Many years of combination of data. They were bought. Quite some time ago, became part of Sir Cana, MPD, IRI, et cetera, et cetera. And the founders have now launched ABCS Insights coming out of stealth.

Karen Lynch: They've been there because they raised a 50 million Series A, not seed.

Lenny Murphy: Series A, I mean, they've been in business, they've been, and they've been successful. Now I've raised 50 million Series A. This is really around causal measurement. So it's, you know, a purchase to exposure. So trying to get the attribution, that was always what Niels Catalina did. Connecting data to go through kind of, you know, not path to purchase in the sense of attribution of, you know, you saw the ad, you went here and you bought the product. Right. Yeah. And connecting that causality. So really cool for a methodology, an approach that has been around for my entire career, coming back with a new approach that's AI-ified around marketing channels. You know, they're dealing with things like being cookie lists and this GDPR compliant.

Karen Lynch: Say, there's privacy constraints being built right into this, right? So there's some shrinkage. Yeah.

Lenny Murphy: Yeah. Or at least a redesigned mousetrap for modern times.

Karen Lynch: For modern times. For modern times and modern kinds of demands, you know, what's being demanded of people right now, which is, you know, data privacy. Yeah.

Lenny Murphy: And I did point out, again, these were not, you know, young, you know, 20-somethings out of Silicon Valley, right? I mean, these are industry veterans, been around all of an age and, you know, doing cool, cool stuff, getting funding, you know, driving value creation. So I love that story as well. So yeah. So there was funding.

Karen Lynch: Yeah, there was funding.

Lenny Murphy: Now, did you list Vendelux? Because I thought this was pretty cool.

Karen Lynch: This was very cool and really, really telling, especially for a company like ours, by the way, that runs events. So, yeah, let's talk about this. Vendelux raised 50 million Series B by Tribeca Ventures Partners. So, I'm not, you know, I don't know this company, but they're expanding an AI-powered B2B event. Intelligence platform. So that's really all I know, kind of is this headline, right? So event intelligence. And obviously, we play in the B2B event space, and they're getting funded to, you know, make sure that there is measurable ROI in the live event category.

Lenny Murphy: Which is hard. Which is hard, right?

Karen Lynch: And I'm thinking big, you know, we're thinking at a much larger scale, big, big tent events. This is really important, you know. I'm not saying it's not important at our own events, but I'm thinking South by Southwest and some of these others, you Dream Force at Salesforce, et cetera, et cetera. You know, these large, large events. Anyway, but isn't it interesting? The market is taking seriously ROI for in-person events. 100%. I mean, that is very interesting to me because conferences in general, at a time when AI is taking over everything, and research technology in our industry, research technology and platforms and virtual. Does give us the ability to do so much online. And now we're really looking to measure the ROI. We're putting funding towards measuring the ROI of these in-person events. They're not going anywhere. And if anything, they're becoming more and more important. Anyway, that was how I read this. I read this and I thought, well, that's really interesting to me.

Lenny Murphy: Well, yes. They didn't get into, at least that I saw, the types of data or data collection, right? But thinking, okay, we've been talking about meta with the glasses and all of that good stuff. Our good friend Emily Fulmer reached out to me a couple of weeks ago. She's talking to a company that has. Like an AI recording device for salespeople for in-person salespeople. Anyway, the point is like there's, there's this, also this component that in live events, technology is getting to the place where we actually could just read AI running in our meetings, you know, to capture data in those environments and help quantify that. And, you know, and Vendelux seems to be. You know, wanting to become kind to the operating system, you know, to do that. So, very cool.

Karen Lynch: Um, yeah, it's cool.

Lenny Murphy: Uh, as you said, the bottom line message is events ain't going anywhere.

Karen Lynch: Yeah, yeah, yeah, I like it.

Lenny Murphy: Let's talk about our friend system one. So, this was an interesting thing. So, I want, you know, no. No secret that I've been good friends with John Kieran forever. I love the Brain Juicer team. I love the system. One team. Susan Griffin came there. So, a special place in my heart. Pay attention to System One.

Karen Lynch: I think we should also say before we share the headline to this story, I view this as. Because I did a deep dive here too, because I'm like, all right, I need some context and I need to fully understand this story. And I view this as a chess game where there are intelligent players, and I'm just going to sit back and watch them play chess. So, like, like, that's how that is the context I want to put out there for what Lenny's about to share with everybody. This is, this is, this is chess. Okay, that's all I'm saying.

Lenny Murphy: This is all I'm saying, my take was: So, there's a large shareholder with Sister Once public company in the UK. There was a large shareholder, Brave Bison.

Karen Lynch: They offered.

Lenny Murphy: Basically a market cap to acquire the entire company. The board said no.

Karen Lynch: And when you first read that, it reads like a rejection. And it's like, it sounds like a drama. Right. And sure, there's probably some drama around it. However, that's why I wanted to give the caveat that this stuff happens.

 Lenny Murphy: And it's happening across. Look, every public company in our space is my position that they are undervalued right now.

Karen Lynch: Every single one.

Lenny Murphy: The NIQ, YouGov, System One, Pure Profile. On the board of a public company that is undervalued. Right. So I have a perspective on this. And here's why. And this is what this was the system, my take on it, the system on board said, no, they're undervaluing the asset, the intelligence asset. And I think that is the story here is the market has not seen yet that these companies, everybody I just mentioned, have a unique data asset that is worth. More than the current and past revenue model attached to it. And so we're in this transformation. And this is an example, right? The companies that have unique data assets are recognizing that that can be monetized in a very different way that is different than a project or a transaction. And that's what they're betting on. Now, the piece is how do you unlock that? That? But I think this is just an emblematic way to watch any of these companies. You've got cool shit. It's worth something more than just how a public market may be valuing it today based on, well, you only made, you only sold $40 million in top line. It's like, yeah, but we can unlock value that adds another $40 million on top of that. It's just compounding. Value of the data asset.

Karen Lynch: And that's the And I think, and that what's happening across the entire marketplace is like AI is making, you know, AI analysis. Everybody can do AI analysis. Like everybody is becoming, everybody's got their widgets, right? And so you need to have that differentiator that what you're doing is really predictive of real-world behavior, or you've closed that say-do gap, or you've done something. That is going to differentiate yourselves from amongst all of the players that have now unlocked the AI benefits. Right. Right. What are you doing? You know, so, so, you know, you've got to prove yourselves now. Like, and I think that that's what's really interesting to me just about this time of disruption in general is that companies, like, I started thinking when I was reading through this brief today, I'm like, you know how, like, when and probably your kids too, because I know you're deeply involved in your kids' education, like they had to show their work. Like, like you couldn't just come up with the answer. You have to show your work. Like, you have to prove that you are smart. So your kids have to show their work. I feel like companies these days prove that you are smart. Like the world has to know that you are intelligent. The pressure is on because everybody has an AI benefit right now. How are you better than everybody else with the AI lift when everybody has superpowers? Everybody has superpowers. So, how are you better? The pressure is on. And you know what? If you are helping companies make decisions, you know, how is your decision making helping them more than somebody else's? Like, you just anyway, pressure's on. So, anyway, sidebar to that, I just know it's absolutely.

Lenny Murphy: I mean, we'll get to that later. Is an ongoing big macro theme, right? Far outside of our industry, but absolutely relevant industry. And I think that example of the compounding, well, let's talk about this new product launch, the stat social AI focus groups. This is a compounding value creation effect off of a unique data asset.

Karen Lynch: Anyway, you want to talk about it since it's well, I just don't know how I feel about this one. This one really, I keep sitting with. So, Stat Social launched AI focus groups using its digital twins to deliver AI-moderated qual feedback tied to its weighted audience data, benchmark validation, all of this great stuff, right? And there's, there's also, you know, they've got like a, people graph, behavioral identities, blah, you know, I don't, I want blah, blah, blah. That's um belittling the technology behind what they have. And I don't mean to do that. So I um, you know, they've got accuracy and they've got scientific measures behind what they're doing. Um, and it sounds smart. So I, again, hats off. Um, so the issue isn't, is it good? The issue for me is we now have AI-moderated qual, but with synthetic respondents. So now we've kind of doubled down AI with AI. And I just don't know. I just don't know how I feel about that because now we've got AI talking to AI. And that I don't know about. That to me feels like, What? Like, I don't even know if that's, I don't know. That's a bigger issue for my head to work through. It feels that that could go wrong pretty quickly. I'd have to see. I'd have to see that. And I've got both of those products. I've been on board with AIQual. Moderated qual, and I board with digital twins. I don't know that they play together in this way. So I just, I don't know, I don't know. I'm having a hard time with it, honey.

Lenny Murphy: So there's something I shared with you earlier this week that we can't talk about publicly, but did you read that?

Karen Lynch: I did Read that. Yeah.

Lenny Murphy: So here's what that makes me think of, right? And besides, guys, sorry, I don't mean to be mysterious, but it was just shared in confidence, but it was like a holy shit uh thing on on how a large brand is utilizing synthetic so we'll leave it at that the now what's interesting is that that synthetic it is grounded on on a recurring constant feed of real human data so I think that makes a big difference here so um uh it's like i've got a pond that's spring fed that's why it's not a stagnant nasty pond because there's a spring that keeps it, you know, flowing. It's the same principle, right? And secondly, and then trained on internal customer data.

Karen Lynch: Yeah. Yeah.

Lenny Murphy: So, in that, if the data sources that are replenishing this, then it's a fully synthetic focus group for some use cases. I think that's absolutely okay, right? Yeah, just throwing stuff against the wall, early stage ideation, yeah. Do you like the blue widget or the green widget?

Karen Lynch: Yeah, yeah.

Lenny Murphy: Uh yeah, Nancy, I can't say which large brand, but brand. Um yeah, but sorry I'll quit my soapbox, but if you dig into that specific example and a lot of other examples that we're hearing about this, the model seems to be looking for low risk decisions. Yeah, yeah, and there's a lot of low risk decisions that we spend a lot of money on the research, so because you still want to cover your butt, you want to have the data, so yeah go there, then have the flywheel of okay now we want to test it against real people. Yeah, and that goes into real research.

Karen Lynch: Absolutely. It's a low investment, but I imagine it's also a low emotional investment. I cannot fathom that buyers are going to be terribly convinced. Buyers are going to buy in. To the results. So I can't imagine it's going to be used for anything that is that serious.

Lenny Murphy: I agree. It's very tactical, a low risk.

Karen Lynch: Our framework.

Lenny Murphy: Elaine Riddell, she frames this up as a consequence.

Karen Lynch: Yeah.

Lenny Murphy: The lower the consequence, the decision, you know, do you like the blue widget or the green widget? You know, a low consequence decision. Should we make a widget at all? That's a big consequence decision.

Karen Lynch: Yeah. Yeah. At Europe, one of our future list honorees who works for the Magnum Ice Cream company, FerranMe, Murana, he talked basically about when is AI good enough, basically, and that every company, his recommendation was that every company kind of have their matrix of is it good enough? Is this a project or a use case when you can use AI for it? And sometimes it's good enough. And other times, like, oh, hell no, it's not good enough for this project. So maybe there are going to be times when AI moderating synthetic samples is good enough.

Lenny Murphy: Well, it's like the sample quality debate, right? We talked about this. Throwing away 20% of your sample or whatever because it's crap or bots in some circumstances, okay, because it doesn't really change the decision, right? I only paid 20 cents each.

Karen Lynch: Yeah.

Lenny Murphy: I mean, there's, there's, by no means am I saying the sample quality is not important. So nobody takes that. Well, then he said we could have a crap sample. No, not at all, but it is through this lens that sometimes good enough is good enough because the decision really isn't that important. So it's a decision that needs to be made. You want to back it up with more than your gut. Fine. Yeah.

Karen Lynch: Okay. Yeah.

Lenny Murphy: But the big stuff, we do more things.

Karen Lynch: So you want to know about that? To Conveo because interestingly, like at the same time when, you when I was like, all right, stat social, you're you're to show me that one, Conveyo launched their kind of next product called Storylines, which is continuous AI moderated video qual, kind of tied to ongoing brand questions. So a continuous Read on consumer sentiment through AI moderated video conversations data. Compounding data, kind of a new kind of brand tracking. Interesting. So always on brand tracking. I thought this was really interesting. And I have to say, I'm apparently a sucker as a writer. I'm apparently a sucker for language like the market is full of plot twists you wouldn't want to miss and continue with. Continuous Read and chapter after chapter. I'm like, okay, I see what you're doing to me. But the idea of a story, right? There is a narrative for your brand, and it is ongoing, and you are then continuously tapped in. So I liked the just like, Just like we've been talking about always on data dashboards, I liked the idea of always on qual. To me, that was cool. I hadn't really seen or heard that it shared in a way that landed so readily, you know, this continuous qualitative feedback. I liked the idea of always on video, basically, where I could get sentiment because they talked about it. Your dashboards show you what's changing, but you don't know, oh, you don't always know why it changed. So it was the age-old, you, you know, your quant gives you the what, and the qual gives you the why. Um, I don't know, it's worth checking out, and I thought this was really interesting. So, this one, like, all right, I see this one. So, um, yeah, I thought that's going beyond the project, right?

Lenny Murphy: Yeah, the uh, you know, it's but it's different than like oh you macro synthesis metasynthesis right it yeah that's there but yeah uh but is that that the always on compounding value of engagement and insights that's yeah you know and to build a data asset you know internally that's just really cool.

Karen Lynch: And also, just imagine the PR benefit for that. Imagine your, you know how quickly things go wrong for a brand when on social media, happens. Like, imagine getting that quick, that quick Read on what's, what, what's happening. I don't know. I just think that like, I don't know, the PR benefit of that sort of thing alone. Imagine the chatter about Taco Bell right now if you're here in the U.S., like, with this, you know, just saying, no lettuce, no lettuce. Um, wouldn't it be cool to be tapped into Taco Bell, like, with a continuous conversation right now? I think that's funny, that would be fun too. Be able to not just the data because the data might be trickling in, you know, but I don't know, the continuous conversation. I think that would be really interesting. And sure enough, a story, right? So, because you can be really responsive if you know why what's happening. So, anyway.

Lenny Murphy: Yeah, that's an explosive story. So, sorry. The safest thing running through my head. There were like five others that would have been a lot worse.

Karen Lynch: I love that. Oh, that was one of the best things you've ever said.

Lenny Murphy: Oh, that bar is not very high. That was it. Oh, gosh.

Karen Lynch: All right. Well, there's no segue to the next story. So we're just going to do it. We're just going to go right there. So the reverse of that. Oh, gosh. Yeah. I know.

Lenny Murphy: I know.

Karen Lynch: We'll get much drier. Oh, my gosh. Okay, okay.

Lenny Murphy: So we both have this. It's the humor of a 12-year-old boy.

Karen Lynch: Oh, my gosh. Apparently, apparently.

Lenny Murphy: This next one. So it was bundled. We'll put it out there. Vox Pot Me, Remesh, Melt Water, all launched MCPs.

Karen Lynch: Big trend.

Lenny Murphy: Cool. But so is that whole other thing. But interesting, two equal platforms, VoxPopMe Remesh, and a social platform, which I kind of think of as quasi-qual in a way. You're listening, you know. So yeah, I'm listening too. Yeah, I mean, please, qualities don't say, don't come after me. It's just a broad, very broad statement of listening versus asking. Point being. I just thought that it was really cool that they were seeing this MCP integration, you know, go where the client is, where the user wants to be, how they access it. But it's not about the survey. It's a different model.

Karen Lynch: Yeah. Well, it's about the accessibility of all of the, of the research, all the research in the world, if it's not accessible. And we've gone from these like knowledge hubs to something else. Like, the greatest knowledge hub in the world is, you know, is now only valuable today if it's built into the workflow with these MCP agents, you know. With the, with the, these insights professionals' ability to get into the transcripts, to get into the survey results, to kind of mine them, to pull, pull, pull from them in that dashboard format or whatever. So activate the insights, right?Make them, make them real and usable. So yeah, hats off to all three of y'all. Absolutely.

Lenny Murphy: Now that's segue in like the big tech news and Karley. Let's talk about the the mark benioff uh quote because that I think that's good place there of uh the currency of ai is trust um salesforce's zero data retention approach this is the continuation of this this conversation happening the last few weeks of big tech companies really kind of take an aim at the uh at the ai the frontier models of saying the days of you get you you getting value from your customers data are over it's the customers data it's not our product and we're really seeing a line being drawn that way and that's as you just said right I think that we're making it accessible there's role for platforms like salesforce and amazon and microsoft and everybody else uh in being the pipes right the connect tissue But shifting away from the compounding value should be the owner of the data, not the owner of the model. And I think that is a sea change. It's a big shift. We've talked about it for years now. Don't get caught in one model. And we're seeing that now play out. So that was really interesting with this emphasis on. The MCP integration makes your data accessible.

Karen Lynch: Yeah.

Lenny Murphy: Don't get caught up in the model.

Karen Lynch: Yeah.

Lenny Murphy: That was my take. What did you have another Read?

Karen Lynch: No, no, just the, just all just all of it, the importance of governance, the importance of everything that we've always said about, the customer owning their own data, everything that you've always said about how important that is, knowing that that is important to you personally. We have, I know, Karley, I don't think you have a pop-up for it, but you know, our AI event is coming up in September, and we are very, very privileged that we have the Director of AI Transformation. Data Insights and User Experience speaking from Salesforce speaking. So I'm so excited for that. She is a star. Melissa Raimi is her name. And I'm so excited to have her. And this just kind of is, anyway, it's going to be an interesting talk to see what she's going to share with us. So just imagine, friends, the director of AI Transformation at Salesforce speaking to us. Yeah. But well, you're. Mr.

Lenny Murphy: Lynch just made a good comment too. All that talk and effort for data lakes is coming into its moment if you put the work into clean, accurate data. Hell yeah. Yeah. Right. We keep coming back to the same, you know, so we talk, oh, so much going on, so much new. But the underlying themes have been very consistent for a very long time.

Karen Lynch: Lenny, we're probably coming up on, we're probably like, it's almost August, which means, have we been doing this two years? Yeah, I guess so. We've been doing this for two years. We've probably been saying the same things for most listeners, is that it?

Lenny Murphy: You're like, God, they are breaking freaking records.

Karen Lynch: This is what they say. They talk about the same five. We probably talk about the same five things. Anyway, it might be true.

Lenny Murphy: It's the fundamentals. The more things change, the more they stay the same, right? I mean, the fundamentals are important. All right.

Karen Lynch: Let's speak of saying the same things, let's talk about anthropology.

Lenny Murphy: Right. They officially launched.

Karen Lynch: They're because I think we did talk about this last week, but they hadn't named it yet, right? Isn't that true? Yes. They hadn't named it.

Lenny Murphy: So they're named. It's Ode.

Karen Lynch: Now it has a name. So. Anthropic Blackstone Hellman Friedman formally launched OD, a 1.5 billion AI implementation venture aimed at embedding AI engineering capacity inside enterprise workflows.

Lenny Murphy: Only 100 deployed engineers? That was shocking to me. That was surprising. You raised 1.5 billion to hire 100 deployed engineers.

Karen Lynch: In one of the press releases or one of the articles, Chris Taylor, who's the CEO of Ode, says, It's pretty easy to imagine this as a trillion-dollar company someday if we execute well. The key challenge of the business is: how do you go through that phase of hyper growth without losing the emphasis on quality? So, don't you think maybe the size has to do with that? Like maintaining quality control and not relinquishing that. If you have a hundred, maybe you still have some control over it. Once you scale up to a thousand or something, you're, I don't know. Maybe.

Lenny Murphy: So, Anthropic, if you're listening, don't get mad at me. I can't, I don't know. A new open source model, Kimmy. Came out today that performs higher than Fable, the so and and it's like 50 cents versus for it's like 50 cents for a million tokens where a million tokens on cloud is 56, so I agree with the thesis. I agree having four deployed engineers going into businesses helping them solve problems and building custom installations for them makes all the sense in the world. I am unconvinced that the current business models of most of the non-open source models are sustainable.

Karen Lynch: Yeah.

Lenny Murphy: The costs, I do that. I love claws. I use it all the time in perplexity. I burn through tokens like that. So I use Grok. DeepMind or whatever. So I'm not spending as much money and I'm getting the same stuff done. My point is, I agree. Quality, smart people doing smart things, the judgment layer, all of that. 100% agree with the model. I don't know that the story is told yet on which of these specific players when dependent on. A financial structure in their core business model, if they're going to succeed. Just putting it out there, guys.

Karen Lynch: Yeah, yeah. I'm pretty sure I read at some point that Mira Marati, the thinking machine, launched something. Isn't it out there now? At least in beta?

Lenny Murphy: I saw something on that that yeah, it's like an implementation. I also think it's like open sourced right or close to open source.

Karen Lynch: Yeah, yeah, so just an interesting time.

Lenny Murphy: Um, yeah, uh but the fact that you know the business model to get people to go into businesses and help AI the hell out of the business 100. Just what's the model? You know, I think that.

Karen Lynch: A tale to be told. And remember, whenever we talk about these kinds of big tech stories, like what does it mean for you? You know, maybe, maybe if you're a small business, maybe if you're, on the supplier side of the industry and you have a smaller company, like maybe you're like, all right, what's the connection here? Just remember, lots of people in our audience work for big enterprises, right? They work on the brand side and they have, you know, there's a pretty good chance that, you know, they're working for PepsiCo, they're working for Mondelez, they're working for, you know, PNG. They've got. Embedded, they've got, they've got AI embedded into their workflows, right? They've got this going across their entire entirety of their organization. So it's not just that they are using research suppliers, it's that their workflows are being influenced by these very things, right? They've got the engineers in their organizations that are changing the way they, the very way they are doing business.

Lenny Murphy: Hence the shift to the NCP.

Karen Lynch: Right? Right. Right. This is the way they work now. So if you are a supplier, you should be paying attention to how they get their jobs done. And if you're not paying attention to how they get their jobs done, then you are behind the times because you need to know everything is changing in their world and you need to understand them, of course. Yes.

Lenny Murphy: Thank you for connecting that dot. That's a great way. And if you're a small business owner and you're trying to think, how do I compete? Understand that, you know, there are. There's lots of options out there you know now as well yeah the uh to get help right um and yada yada all right this uh um I don't know what to make of this one it was kind of an odd duck but I thought we had to put it in here um the video amp in nielsen one quit mrc accreditation and Com score the only accredited TV currency. For years, you look at the, yeah, that there's a whole sub-industry around accreditation of media measurement. So MRC, the Media, Media Research Council, I think that's it right.

Karen Lynch: Yep, yeah, media research council.

Lenny Murphy: Your rating council, media rating. Rating council, rating council. Thank you. They've owned that and companies had to, you know, pass their their uh their accreditation kind of like iso certification type of thing and now I have two of the biggest players and one the the biggest player nielsen yeah yeah saying yeah we're not gonna do this anymore yeah that I don't know what message that is I don't know I think um I don't know I think um are they are they trusted measures are they um credible measures?

Karen Lynch: Did something, are they, are they validated measures? Like, I don't know them well enough, but, but, you know, it's funny because I think I told you that my, you know, my bonus son has that Nielsen was here and he had that Nielsen thing in his pocket or whatever. And, um, he was really taking it. He was really taking it quite seriously. Like he made me bring his charger, was staying here to recover. He was like, don't forget my charger. Like, so I had to bring his, you know, he charges it. And he, he, he, like, took it really seriously and stuff. But also a world where somebody made a joke at some point, like, couldn't somebody else, you know, couldn't you want somebody else? Like, couldn't Tim put it in his pocket or something, like at any given point in time. Like, is it really, is it accurate? You know, like there was this, there was this kind of joke, lighthearted joke made about it. He was going somewhere, and or, you know, do you want us to leave it at my parents' house or something? Ha ha ha, because I watch different shows. Like, you know, like the GameShare Network.

Lenny Murphy: Boy, he spent lots of time on the GameShare Network.

Karen Lynch: Anyway, so no, but for real. So I don't know. So, I don't know. I don't know what the story is there. But maybe there's something there. And it requires a deeper dive, perhaps. It does.

Lenny Murphy: It's interesting because it comes up around like, synthetic. Almost every industry has some type of better or good housekeeping seal of approval, some type of ratings and review certification model. So you're not getting sink oil. Um, yeah, at the same time, any company that subjects themselves to that is losing uh IP and the opportunity for innovation. So, I've always been kind of of two minds about that. You know, why should I do what you say I should do? Because I think this is better, right? I want to do this. Um, yeah, so it's just interesting. And all that in the era of AI, and I think that's probably what this is about. They're using new data sources, new approaches to measure this. Maybe the MRC is not caught up and they're saying, we're not going to wait on you. Or let the market speak. I don't know. I'm just guessing, but it's an interesting development.

Karen Lynch: Yeah, we're guessing. So if you know, email us at theexchange at greenbook.org and educate us because we're just guessing. So, yeah. All right. All right. Oh my gosh, it's like almost a quarter of it, and we still have like three. We have three kinds of special interest stories. So let's rapid fire these and let people get on days, shall we? Okay, yes, ma'am. Strap seven, synthetic data study. So this is worth a read. Download it, check it out. But they found that it's a research, you know, a research paper, found that synthetic. Survey data is improving at top line replication, but still breaks on change over time. Tracking change over time is failing a bit, but improving over time.

Lenny Murphy: Yeah. So yesterday's date is not necessarily predictive of tomorrow's outcomes.

Karen Lynch: Yeah. Yeah.

Lenny Murphy: So definitely worth reading. This one. I think we need to watch this. Uh, this came out of the blue. Um, Zaley Global securities fraud lawsuit over opt-in data claims.

Karen Lynch: Yeah, yeah.

Lenny Murphy: This they said that uh they had 240 million opted in individuals, yeah, and they're being half of that for their data source, and they're being sued. Yeah, um, so panels. Just saying when you say we got a million people and it's really like, 100,000 actively participating, you may want to be careful about that.

Karen Lynch: Yeah, yeah, panel companies or anybody with that opt-in clause for that research, just check out the ruling. I mean, isn't that just fair to say? Check out the ruling, do some research.

Lenny Murphy: Yeah, because has it gone to court yet? It was a motion to dismiss and they said, nope.

Karen Lynch: Yeah, we're not dismissing it.

Lenny Murphy: And this is Southern District New York. This is this way, this is you know, this is New York suing, yeah.

Karen Lynch: Yeah, so uh, yeah, pay attention guys 240 million opted-in individuals, so we're not talking about a small-scale situation here.

Lenny Murphy: Nope, uh, and this was. Yeah, I'm not sure of all of the details on that, but SATA does record data. You know, it's advertising, you know, measurement. They pay attention, guys. Yeah. Yeah.

Karen Lynch: All right. Last one. Last one.

Lenny Murphy: Another strat seven. Well, connected to the strat seven.

Karen Lynch: Connected. Yeah. So this is published in, you know, the MIT Sloan Review. And I actually shared that on LinkedIn. Also, it's a really cool study. Some great people are cited in it. It's reviewing where Gen AI can and cannot support consumer insights work. And it reinforces the distinction between AI as a workflow accelerator and AI as kind of a source of consumer truth. So where it's great, where it fails. It's great at synthesis, summarization, pattern recognition, but maybe not so great at complex behavioral segmentation or audience understanding what's current, what's now. So there's just digging in, checking it out. Do your own thinking about it. Anyway, it's, you know, AI is an accelerator of some insights works for sure, but it is not the, it's not the end all for every project. So, um, which I say, I think right, all of us have netted out. So, um, you know, but good study and, you know, good publication there.

Lenny Murphy: Yeah, I think the MIT Sloan is actually exploring this, right?

Karen Lynch: Studying this. So it's great. Yeah, very cool. Very cool.

Lenny Murphy: All right. Well, happy Friday.

Karen Lynch: Happy Friday. Yes. Yeah, I know. Thank you. Yes. You have a good weekend. Also, for everybody in the Northeast, I think I hope that your day is cooler than it has been, that the air is fresher than it has been, because we've been dealing with the heat wave and the Canadian wildfires. It's been like, it's been bad up here. So I hope for those things. Up for everybody up here. I don't know. How's it been down there? Have you been getting smoke from the fires down there?

Lenny Murphy: No. All of our weather patterns tend to come from the west, the south and the west. We don't get much from the north.

Karen Lynch: We have had a bad few days.

Lenny Murphy: So I've heard. Yeah. So everybody, yeah. Don't. I'm a smoker, you Uh, so I'm used to coughing my head off all the time.

Karen Lynch: Oh my gosh, Lenny, Lenny, Lenny. Yeah, all right. Well, I wish, I wish you would knock that off, but you know, what are you gonna say?

Lenny Murphy: Other conversations, it is my only flaw. That's the one that's.

Karen Lynch: That's what you're going with.

Lenny Murphy: Okay, that's fine. I am perfect in every other way.

Karen Lynch: So goodness gracious, friends. We'll end the show on that note. And Lenny and I are about to get on a Slack call when I'll say, seriously, that's what you want. My friends, good to talk with you all. Lenny, I'll talk to you in a few minutes via Slack. Let's take a little break and we'll see you all next week. Yep.

Lenny Murphy: Thanks, everybody. Take care. Bye-bye. Bye.

Karen Lynch: Yeah.

Links from the episode:

NCSolutions Execs Reunited at ABCS Insights 

Vendelux Raises $50 Million Series B To Scale AI-Powered B2B Event Intelligence 

System1 rejects Brave Bison takeover bid  

StatSocial Launches AI Focus Groups 

Conveo StoryLines: A continuous read on your consumers, chapter after chapter. 

Remesh Announces the Launch of Remesh Connect

Voxpopme Adds GenAI Access to Survey Data via MCP 

‘Almost unlimited’: Execs says AI demand remains strong even as enterprises move to ‘valuemaxxing’

Anthropic, Blackstone bet the next trillion-dollar AI business is implementation, not just models 

VideoAmp, Nielsen One Pull Out Of MRC Accreditation Process 

STRAT7 study finds synthetic data has improved, but still falls short beyond topline results 

Investors Can Proceed With Suit Against Zeta Global, Judge Rules 

How GenAI Can and Can’t Help Manage Customer Insights