The Exchange

November 19, 2026

20+ min read

What Happens When AI Agents Start Spending Your Budget?

Who owns the customer relationship when AI does the buying? Explore AI shopping agents, synthetic research, and the future of marketing.

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!

Who owns the customer relationship when AI does the buying? From big acquisitions to interview-based twins to agent-managed budgets, in episode 153 of The Exchange, Karen Lynch and Lenny Murphy dig into the questions research and marketing teams can’t put off. They also cover where synthetic research holds up, where it doesn’t, and why the answer comes down to the quality of the human data underneath.

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

IIEX Europe 

IIEX Asia Pacific 

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

Transcript

 Lenny Murphy: It is showtime. Showtime. Happy Friday, Karen. Happy Friday, happy Friday, everyone.

Karen Lynch: Happy Friday.

Lenny Murphy: We were just talking. You've got a busy week coming up. Well, I mean, too busy, right? But now we're here.

Karen Lynch: This week hit hard for most of the team here at Greenbook. I don't know what we expected, but everybody I'm talking to, it's like, whew, a busy week. You know, because IIEXWest is a new event for us, we did not, we didn't know what to expect in terms of equating it to our other events. I think we thought it would be as small as potentially as small as APAC, which feels like a low lift. We didn't expect it to become as big as. Europe, or at least start to head in that direction. So, as the registration numbers started to go up, so did our cortisol levels. So, yeah, so energy excitement is going up. You know, stress is going up. We look at this amazing graphic, Karley. Thank you. Like, you know, I have done the math. 60% of the speaking spots are featuring brands. It's an extraordinary amount of brands that are coming to this, not just on stage, but as I look at the registration list, they are showing up for this one, and it is extraordinary to see. And I just cannot wait across all categories. People whose names are new to me, people that don't. Necessarily come to our North America event? Um, which is why we wanted to come to them. You know, we wanted to come out because this was our whole intention: let's bring our event to them because they are not able to come to us for whatever reason. And um, and yeah, and so uh so yeah, next week's going to be a ride, and I'm really excited.

Lenny Murphy: Very cool. I am excited as well, and then so uh you. You'll be traveling next Friday. Yes. But we have a historic merger happening next week, right?

Karen Lynch: It's going to be awesome. I'm actually quite jealous. Will be in the air and not even be able to, I don't know that I'll be able to listen live on air because I don't know that I will have the Wi-Fi to do it. But if there is a way that I can listen in, I will do my best because it's going to be great. Yeah.

Lenny Murphy: Well, let's just get the shout out, right? So, um, Matt Valley, the rock and roll, we're bringing the rock and roll to The Exchange.

Karen Lynch: The rock and roll to The Exchange. Yeah, you know what? He is, you know, a long time listener, first time co-host. Um, and I think that he um, he will. He will do justice to this stint as co-host. He will also be in the West. So that's one of the reasons why he, I mean, so are other folks that have stood in for us before, but he will be chairing some of the tracks. And I know that he will be actively engaged with the content. He's actually chairing the track where the competition is taking place. So he will be particularly primed to. Share how the competition went, which I think is a really important thing to debrief with you as well. So, yeah, so he was very top of mind when it came to who could do this with you. And so, not only is he going to be, you know, good at the show content, but he'll be able to talk about the event and, in particular, the competition winners. And so, really excited for him and excited for you too. Yeah. Ah, there he is.

Lenny Murphy: All right, Matt. Yeah, there we go. All right, we're throwing the horse. So, the uh, that's great, that's great. Yeah, looking forward to it. It's gonna be a lot of fun. Um, it will be, it will be, all right, but sweet and busy, but my god, there's 80 different articles that we were in the consideration set for this week, 80, um, which. I don't know if that's a record, but it seemed like Lord. And every time, it seemed pretty close.

Karen Lynch: Yeah. And every time, you know, every time something went in there, and, Lenny and I both have our processes for scaling them back. You there's a lot of discernment that goes into this and about what, you know, what we pull out. You know, it's not necessarily an easy thing to make these decisions. Because we could really take this show in many many different directions so um you know there's a lot happening so we really we really do try to focus and um keep this show to uh you know 45 minutes remember remember the days we thought it could be drunk 50 or 20 right right we quickly learned it well there's I blame that I blame the volume of content, not us, on our ability to maintain that. But I will say, you know what, I didn't even have a chance to talk to you about this, but I was talking to somebody recently and they were talking about kind of the premise of the show. And I was like, maybe it's good that I take a moment and tell, say the premise of the show and kind of the backstory of the show for new listeners that don't know. Lenny and I, we, we, we talk periodically, or talking periodically. And now, what has it been two and a half years ago? Like, I'm losing track of how long ago it was. And we were like, you know, we should maybe do this. We should maybe take this on the road. This idea of talking about the industry with each other, but live in front of an audience. So this show is actually designed to be a show where Lenny and I talk about the industry, but let other people listen in on our conversation. Like, really. That's kind of what we're here to do, share out with you a conversation that we may be really having with each other and sharing out the news. Yeah. That's really how it all began. And we thought maybe it would be 20 minutes. Do you remember what it was like, like what we talked about, how this first came about?

Lenny Murphy: Yes, yes, absolutely. And we, because, and we thought, let's, because Grimmett doesn't do a news thing, not like MR Web and Research Live, and you know, but we certainly pay attention to. All of those things because it's relevant to all of our business interests. And to contextualize that, separate the signal from the noise. And I still think that's what we do back to the original point: there's so much signal. There's a lot of noise, but it's also a lot of signal. Too much for most people to keep track of And so our job here is to. Try and convince the most prominent signal, I think, is the way I begin to think about it. Is that we may have 20 articles or 20 submissions that are really talking about the same issue in different ways. And to bring that to you, our dear and wonderful listeners. Hopefully in an entertaining and contextualized way.

Karen Lynch: Yeah. And allow you to do some critical thinking too, right? You, the listeners, do some critical thinking too, because you know, we're connecting dots for you, but you're, you know, you're then filling in the picture yourself. So, um, you know, take what we're sharing out and talk to each other about, and then, you know, share it with all of you, our listeners. So anyways, just so you understand that that is sort of what the show is about. It's Lenny and I talking to each other at the same time that we're talking to you, um, and making sense of all of it.

Lenny Murphy: So, yep, yep, that's it.

Karen Lynch: So, let's start, let's start with what Salesforce is doing because we need to make sense of what Salesforce is doing because they are. If you want to think about the world of signals through the noise, they're being very loud right now.

Lenny Murphy: Very loud.

Karen Lynch: Yes. They're shouting at us and you may not understand what they're saying because you're quivering because they're being so loud. So yeah, so Salesforce agreed to buy a company called Braze for $5.66 billion this week. And if you remember, last week we talked about how they bought loops. And of course, just a few weeks ago, they bought Listen Labs. They are lining up AI-run customer interviews, AI explanations of why metrics are moving, real-time customer messaging. You know, in September, I don't think we talked about it, but they bought a platform called Contentful, which I pay attention to because it's a content platform. But I'm like, ah, the industry doesn't need to know they bought contentful. So I think that was one of those stories we scraped because who cares that they bought a content platform? However, looking back, it's just another acquisition that they are integrating into their offerings. And that's what makes us pause, right? Like, okay, certainly listening was a story, and then loops was a story. But now with this braise, it's like, okay, they are moving in a very specific direction. And that's what we're, that's, that's now what we're putting a pin in for everybody. Like, take a look at what they're doing here. This is, they are broadening what they are doing well beyond what we've known them for.

Lenny Murphy: Absolutely. I think the way I sum it up. Up in the top, there's always conversations with other folks all through the week and Salesforce continues to be a, oh, to your point, they're loud, they're building the workflow infrastructure, the way I think about it, and the stack, right? All of these acquisitions reinforce different elements of the workflow. To be the infrastructure for how an enterprise manages everything related to customer information, right? I've always thought about the marketing life cycle as insight, sorry, to engage, understand, and activate. That's just been my general view, right? Everything kind of hits underneath that. They're building that on steroids.

Karen Lynch: Yeah.

Lenny Murphy: With all types of resources that fit underneath those big banners. And think about it from a disruption standpoint. It's the only thing that these companies can do. Because, and sorry, but we should point this out, they're also one of the first platforms to say, we don't own the eyeballs anymore, we don't own the interface. That's not the deal. So they gave that up to willingly say, this is going to be embedded into whatever AI systems you're utilizing from an access point standpoint. But everything happens under the hood, right?

Karen Lynch: Yeah.

Lenny Murphy: They own the process. And I just think that's incredibly telling. Anybody else, right? If you think that you own a workflow or a piece of the workflow, there's a difference between owning the eyeball and owning the process. And Salesforce is absolutely in it to win the process.

Karen Lynch: Right. Because at the end of the day, you know, at the end of the day, everybody's working for the sale. So they're talking about: we want to be involved with how you talk to your customers, the messaging component, how you communicate to them. We want to be involved in how you understand them, the CX component, how you understand the analytics, the shopper intelligence. We want a piece of all of these different businesses. Not just the CRM business that you used to rely on us for to be a database of your, of your sit for in your sales department. Like, nope, like that's not it. So, anyway, they have a stake in a lot of different industries all of a sudden. And I think that's really interesting.

Lenny Murphy: It is. Somebody asked me who was going to, who is next doing that. And it's like, well, who are the usual suspects, right? That Oracle, SAP. Adobe, Microsoft. I mean, those will, they all obviously are in a position to want to own more of the stack. I heard this week that some are big, so I'm navigating NDAs, several big tech companies. That you wouldn't normally think of as playing in this space are reaching out to companies in this space to build their version of the stack, right? And fundamentally, it has to do with data that drives greater optimization across their core businesses. So I think we're going to see a whole lot more of this and from the usual ends, probably some unusual suspects.

Karen Lynch: Yeah, yeah. Well, speaking of unusual, unusual subset, subsets, yeah, yeah, suspects. So, this company called Qualitate, which popped out of nowhere. Um, I'm like, okay, qualitate. I just wanted to get into the game, I suppose. They were like, stop talking about all these other players. They raised 25 million Series A, this company called Qualitate, led by Insights Partners. This was just this week for an AI interviewer platform running thousands of quality or proactive expert conversations each month, then turning them into a searchable research library. I don't know if that sounds familiar to you. You know, fascinating. So I'm like, okay. Again, this didn't seem like a new company and they're saying like they're the first continuous intelligence company. So I had to dig a little deeper to figure out why. But, you know, they're targeting B2B buyers, right? And industry experts. So therefore they might be trained differently, right? Whereas I think everything we've seen with, you know, with Listen and Outset and Conveyo has been very very consumer targeted, right? With a lot of the brands. And that's why it pops up, you why they come, you know, first to market. But it looks like Quality's going after a B2B audit, B2B audience. So the questions might be different. So the AI moderation training might be different. So I was like, okay, interesting. So they are definitely getting in the game. And sure enough, getting 25 million Series A I'm like, here we go. Anyway, I just thought it was really interesting. And you know, you think maybe money's running out for these guys, but it's not.

Lenny Murphy: Well, let's kind of talk about this, so Karley, sorry, we're going to the GLG. Exciting, it's related, right? The GLD, GLG, Gershman Laming Group, and expert network. Uh, you know, I get pinged by them all the time to do stuff. And the, uh, uh, frontier giving AI labs verified expertise to evaluate fine-tune stress test benchmark models. But recruiting and vetting people, that's the core thing. And we've seen that theme over the past few weeks specifically: of new panels for all intents and purposes, right? Standing up related to experts. And now, this, you know. The qualitate, I think, sits on top of that. It's like, you've got all these, you know, experts. Now let's turn that into an ongoing stream. That's really interesting. And I, how much you want to bet that their clients want that for agentic training.

Karen Lynch: Yeah.

Lenny Murphy: I will almost guarantee you that is a primary use case. And I have no, you know. Or some version of synthetic.

Karen Lynch: Yeah. Yeah.

Lenny Murphy: But always, and if you're an expert, I'm sure you've done too, right? Expert interviews. That's, it's, you know, that's an ad hoc. It's expensive.

Karen Lynch: Yes. Right? Very expensive and very different. They have a completely different feel. The questions by design are more. I'm even slowing down as to you, deliberate, deliberate in the way you ask them. They're much more often, they have to be more intelligent questions because the vocabulary is often different. If I were talking to dermatologists, I did medical work, I would have to ask very, you know, questions that just aren't everyday conversation. We're not talking about person-to-person daily wants and needs. We're talking about medical wants and needs, or industrial wants and needs, or even just corporate wants and needs. We're talking very differently. So, the AI would have to be trained extremely differently, and each industry is different. So, it's very vast and more in-depth. So, anyway, yeah, so back to following the money, you know. Money to be out there too.

Lenny Murphy: Yeah, scaling access to humans, yeah, you know, across all of these different use cases, right? Yeah, that's the deal. We've been, I don't know about you, but I keep every week now, and I'm cautious of not drinking my own Kool-Aid, right? But I stress test and kind of, red team my own thinking on a regular basis. And I still could come to the conclusion that fundamentally data.

Karen Lynch: Yeah.

Lenny Murphy: And what the hell do we do with it? And that's where the value compounds. And now everything else, this is an example, is scaling that process to get the right information to drive the right decision. Yeah. Yeah. So. Anyway, uh, and thank you, Carla, GLG.

Karen Lynch: Yes, I moved it up. I'm sorry. Oh, you did. So I'm multitasking. I'm multitasking. I couldn't stand it because I was like, I wrote that down there. Let me screw this up. My brain is in that state of mind. There's a problem this week. I'm fixing it because that's where I am. West has me like fixing problems every five minutes. Um, yeah, I was just gonna say something else about um. This situation. If there is somebody out there who has an AI-moderated QUAL platform that is targeting the healthcare industry, I just started to think about how hard it is to conduct qualitative research with doctors as I was thinking about my dermatologist situation. Like, it is so hard to schedule those interviewers. I mean, those interviews are because of a doctor's schedule. And the coordination of all that. And I started to think just about the use case for medical research. And I don't think any of these companies, I'm sure they're capable, and I'm sure they could all say Karen, we do medical research, but one that is specific for that, we haven't seen that yet, have we, Lenny?

Lenny Murphy: I think Canovo is doing that. Okay, okay. Yeah, because they bought, they bought a doctor, not a doctor directory. They've bought Rare Patient Voice. Yeah. They had InCrowd, and InCrowd was kind of a quant element of that. Yeah. So I don't know for certain. I can't point to a news item off the top of my head. So if they have it, it's probably someone like that or IQV or, you know, that's, because I agree with you. I mean, building that specialty capability 100%. Yeah, I suspect it would be somebody who's already a major player in healthcare insights.

Karen Lynch: If they haven't done it, they're about to.

Lenny Murphy: And if you're listening from IQV or anybody else, let us know.

Karen Lynch: Yes, let us know. Shoot us that email, The Exchange at greenbook.org and say, talk about us next week or two weeks from now when I'm back. Yeah. Yeah. Okay. All right. Let's get into Brave Bison, man.

Lenny Murphy: Because the soap opera.

Karen Lynch: The soap opera continues. Yeah. You know what? You know, go on, get into this. I updated this particular bullet point a million times and bounced back and forth from yours to, you know, there's it's this takeover and the, you they're falling short of what all the stakeholders can agree upon or the shareholders can agree upon. And the bottom line is system one's independence is on the line here. You know, it's just a matter of time. They're just you know, is it inevitable? They're not going to agree. When's it going to happen? You know, system one will be a part of Brave Bison.

Lenny Murphy: It's just at some point they buy the shares or something, right? They're at 35.2% already ownership. Yeah, it's interesting. And the most, so we'll see how that plays out. They have a new date. Uh, the 26th, 1 p.m. London time on October 26th. Um, yeah. Uh, but the uh they want that data, yeah, that's what they want. They want to unlock the access to that data across their, you know, their uh, their ad advertising uh systems, yeah. So uh, and obviously. System one knows they want that data. They know that it's valuable and they keep going. No, no, I'm sorry. That's not enough. So, we'll see.

Karen Lynch: You know, as I think about it, isn't there a benefit to system one about having bigger money behind them? I mean, this is not, I mean, I understand, you know. I understand wanting to be independent, but I understand wanting to detach the marketing company from the ad testing. Like, I get all the reasons why they don't want this, but also, isn't there a win-win at the end of having more money behind the work that they do?

Lenny Murphy: Like, I, I don't know any public, well, that's not true.  As chairman of a public company, you know, chairman of the board of a public company, being a public company is a pain in the ass. So it truly is. There's benefits, and you know, obviously, but significant challenges and complexities in doing that. So, my default in this scenario is like, why would you not want to go private? It's so much easier. And I know a lot, you have quite a few other public companies in this space that I know those CEOs feel exactly the same way of like, oh my God, can we just go private?

Karen Lynch: Yeah.

Lenny Murphy: You know, but they have shareholders and shareholders have expectations on the return. And we'll see. Not many public companies in this space are doing so. Or are doing particularly well from a valuation standpoint. Doing well, their financial performance is just fine, but the market is undervaluing them. Yeah. It seems like system one is banking that will change. Yeah. I hope that it does. I'm not seeing, there are several public companies that I'm close to the CEOs, and every conversation is. The market doesn't understand us and doesn't credit us appropriately for what we have. I can't disagree with them.

Karen Lynch: So I don't know. Here's what I would think. And this may be me, I'm jumping to a lot of conclusions here, but in the ad testing business, right now, with all of these AI platforms coming at them with AI tools. Ad testing is eating up market share in that business. I would think their business, if it's not suffering already, is not growing right now. And maybe, maybe, maybe I'm not, I mean, I'm not an expert in this field, but I see the potential for these AI-moderated ad testing platforms to just continue to grow. I mean, even. Ad Week last week, when I was watching the electric twin demo and I saw how they're using synthetic twins for early ad testing, and I'm thinking, why wouldn't I just do that for early, you know, early ad testing? Why wouldn't I if I had a company and I wanted to just do some quick ad testing? I could not come up with a barrier to use. I couldn't do it in my head. So, if I'm a, if I'm a company, if I'm an independent ad testing firm and I am seeing the future with all of these other AI native ad testing platforms coming at me, why am I being so bullish right now? I don't know when. Again, not an expert.

Lenny Murphy: The data will drop. We're about to talk about our friends while listening, you again in a minute. Because to your point, the synthetic, yes, that is a perfect use case for early stage ideation and testing. 100%. And then you go forth and you validate. Now, a company like System One or anybody else that's sitting on a proprietary data asset that is effectively a normative benchmark. That's valuable. That drives those things. And there was, I mean, we had so many papers about synthetic we included, and we chose not to.

Karen Lynch: So many. Yeah. So many.

Lenny Murphy: All come to the same conclusion: you know, that they're only so good because of the models. The only thing that makes them better is human, ongoing, consistent human data that calibrates them. So it's just a calculation, I guess, on where that value is to be created.

Karen Lynch: And they're betting on the value of that data. And to your point, that's what, you know, and I think nobody is saying that that data isn't valuable, and that's what they're holding on to. So, you know, I just know that that competition where we do have expertise, you know, if it's not in the wheelings and dealings of takeovers and we do have expertise into what into the platforms that are coming. We can see the future of this industry, and we know it's more we're not seeing less of the AI accessing platforms. That is where we are, experts.

Lenny Murphy: Yes, 100%. And it goes back to the big tech companies, yeah, who are you know, mostly. Largely advertisers right a huge revenue stream for all of these companies and many their only revenue stream or their you know is is advertising they want access to human data so they can optimize the advertising same old same old the uh all right all right yeah speaking of twins twin talk but you know listen listen right listen lads man now I gotta say saw this coming saw this coming it was the logical extension for the it was logical so I'm on I'm on um I'm on with uh with with uh bridget yesterday and we were looking up something on linkedin and I I stopped her in her tracks and I said hold up alfred just shared

Karen Lynch: A video and um we stopped, we stopped what we were doing stopped in our tracks to see you know to watch alfred's video um listen labs if you didn't see it all happen yesterday, Listen Labs launched listen twins. Each twin built from an in-depth interview with a real person.

Lenny Murphy: Yes, that's a real twin. We can call, we absolutely can call this a twin.

Karen Lynch: This is a twin, and the twin will be continuously fed with, you know, so it's going to be one of those kinds of simultaneously updated digital twins so that you can. They'll continue their live interviewing with the person. At the same time, their twin can be accessed for other use purposes. Fascinating. Yeah. So, you know, two uses all baked into the Listen Lab platform. Whether you want to go live, whether you want to go with the twin, they've got both.

Lenny Murphy: Expanding cases. Right? The uh uh which and you wonder why Salesforce bought them, right? The and they're they're expanding the I do think it's this thing you mentioned, and we don't have to get too far down this, but it did make me think of what is that at some point. The industry is going to have to tackle head-on this idea from the respondent of wait, you paid me 10 bucks. To do this interview, but now you're making, you know, right, whatever, you know, $10,000 off of that.

Karen Lynch: Yeah. Yeah.

Lenny Murphy: There's going to be something that we're going to have to deal with. And there was a couple weeks ago, I remember that company was sign shift, signal shift. I always forget the name. Yeah. We talked about it. And that was their play, right? Was to get a consumer say, wait a minute, you should get a piece of that. So there's broader implications. Hats off to listen, logical thing to do. Fantastic.

Karen Lynch: Absolutely.

Lenny Murphy: As an industry, we're still going to have to think through what this means to the respondents when we are monetizing them on an ongoing basis rather than a one-off survey. That was one thing, right? You pay value, one-off survey. But now we're creating products.

Karen Lynch: Yeah. Well, there may be something that, you know, again, next week, I'll be paying attention to who's talking about this. There may be room. In the QA portion of some of these talks or even some of these demos. Know. I don't know who's demonstrating what aspect of their platform next week, but there may be room for that kind of a question: how are respondents getting compensated when their data is being used to train a twin? How are we doing that?

Lenny Murphy: Because musicians. Right? I mean, you know, you create some, you, whatever, and you get a royalty going forward. Anyway, you know, those types of things are near and dear to my heart.

Karen Lynch: Yeah. Yeah.

Lenny Murphy: On a related note, we should be another competition winner.

Karen Lynch: Right? You know what? Both Alfred and Dana Kim are going to be sitting on this panel next week. I can certainly say to them, not during the panel, because that's not on topic, but afterwards, I can pull them aside and be like, hey, friends, let's talk about this. Dana Kim from Highlight Man launched Pulse with twins modeled on individual at-home product testing. How sweet is that? I was like, okay, that's cool. So, yeah, pretty sweet.

Lenny Murphy: And there's a big thing on this, too, about this synthetic sprawl category. Proof of provenance and tracing. And both of these solutions can be traced to a specific respondent. And so there's a path of evidence there. I think that's really important. It's going to grow in importance. So hats off to everybody who's doing that because you're creating a record. You're creating a traceable record. And that's important for GDPR in California, a lot of good stuff too. But it's also just important for confidence building. So, you know, somebody, if there's a digital twin of you or I, be able to validate, okay, that really is, you know, Lenny and Karen, they're, you know. Middle-aged mom, middle-aged dad, right? Whatever their body behavior, we're using it to base a billion-dollar decision off of And that's the key. I want to make sure that if that billion-dollar decision, it's a CYA, right? So it's just cover your ass that it's not hallucinatory bullshit, that it actually is based on real data. And these, you know, they're, they're, they're building that.

Karen Lynch: Uh, yeah, that's important too. Yeah. Yeah.

Lenny Murphy: Uh, The circana, interesting synthesis. They're having weekly item level data from online retailers starting in 2027. So that's cool. All those big companies that are already industry standards on data, they, yeah, and IQ, since we didn't have an IQ. Exactly.

Karen Lynch: We have to mention them this week.

Lenny Murphy: It doesn't seem like an exchange unless we mention them.

Karen Lynch: I don't know if it's because the NIQ didn't ship anything this week, or if we became, what do you call it, like tone deaf, or if we just, you know, our eyes glossed over it because we had had enough.

Lenny Murphy: So I think they're just selling. I think they went through their shipping period, and now they're just like, oh, we got to sell all this stuff. Um, yeah, but don't count out, not that we've ever counted Circana, big company, right? But these companies continue to make moves as well in consolidating and synthesizing information to the same thing, infrastructure. They want to be infrastructure.

Karen Lynch: So, yeah, yeah. Uh, let's talk about another. Oh, yeah, you know what? I should have bumped this one up because it's yet another one of our competition winners that also shipped a new product this week. So, Ange from Ad Maple, again, like now I'm realizing I'm out of order here. So, this spring's competition winner, Ad Maple's her company. She shipped AdMaple Mosaic. So, this was not, this just shipped this week. So, these are some LinkedIn posts. Alfred's was a LinkedIn post I think we're sharing, and then we'll share Ange's LinkedIn post explaining it. This is a statistical engine that is always on kind of proactively investigating your data. So, you know, while your data is there, it's going through and looking for insights on your behalf, really. It's looking across your surveys and your trackers and your interviews, you know, built to flag findings that you may miss, things that are going to move your business that you may miss. So proactive, and it's not AI, it's statistical analysis. It's a statistical engine that is at work. Super interesting, you know, complex, way above my head. There's math in there. There's math and engineering that's way above my head. That's one thing about Ant's platform in general that's just, you know. This is an engineering-based way above my head. So, hats off to you, Ange. And she will also be at West demonstrate, she's got a demo and a kiosk as well, demonstrating her product. So, our winners are representing next week. It's really exciting.

Lenny Murphy: It is, it's really cool. I wish to point out that, you know, so I'm not sure how big Ange is now as far as the team, but I mean, we're talking about, you know, they're doing a lot of stuff with very few people. Yeah. So, I, you know. Don't know if it's two or five or whatever, but it's not many.

Karen Lynch: Yeah.

Lenny Murphy: But boy, they're leveraging the hell out of these tools to keep shipping. So I have been obsessed for the last few weeks with this, with this agentic commerce issue and the implications of that. You probably saw that in the court.

Karen Lynch: Before you speak, I want to make this very clear to everybody listening. If you do nothing after this show, watch the MasterCard video we are about to talk about. Yeah. Yeah. That's it.

Lenny Murphy: This, so I'm sure you picked up on this, but just for our audience, the shift towards genetic buying, especially on the consumer side, fundamentally changes everything about branding, advertising, and the path to purchase. These companies like Circana, for instance, right? I am quite certain that they are trying to figure out how to capture the same level of data that they have for consumers when the agent is doing the buying. And that, there is a fight over this whole thing, way above our pay grade on who owns the data, who owns the customer, who owns the customer's data, and who owns a transaction. And there's also implications, you know, all this other stuff. Anyway, I'm just fascinated by this because there are lines and companies like, no, we're never doing it. And there's like, well, hell, we're all in. You know, and our first article was, you know, Walmart's one of the companies like, we're all in. So they opened up to shopping agents with profiles. And that's why it said Shopify, but Amazon, like, oh, hell no. No, no, no, no. And many other brands. Although my cynical view is that you're not worried about fraud. You just, you want to own the data. You need to understand who's buying and you can't do that with the agent. And it's just so interesting. Um, and then the master you were the master card video, obviously.

Karen Lynch: Again, so so while Walmart is doing what they're doing, and yeah, Shopify is doing what they're doing, and you know, Amazon's not doing what they're not doing, the MasterCard demonstrates how they are letting the agentic purchasing happen, like how they're saying, All right, retailers. Do what you will, but here's how. And they have an example of: we're going to let you set a budget and we're going to let you decide how you want it to work. Do you want it to be kind of like you establish a budget, and here's how we can basically manage your budget for you. We'll put this much towards this, this much towards this, this much towards this. This was an example of a campaign. So, this much towards video, this much towards audio, this much towards, you know, creation. And it managed, the agent managed the budget that you set aside for creativity. It, the agent, managed a set. The budget that you gave it, it means master cards agent, including buying research, including buying research. And I just watched it and sat back and thought about that as a game changer. The exact thing that we've been talking about now: with that changes everything, that changes everything. Could you imagine? Because, because. There wasn't a person that approved those purchases.

Lenny Murphy: Nope.

Karen Lynch: The agent just paid for it. Said, okay.

Lenny Murphy: I had a moment earlier this week. It was late in the evening. It had been a really busy day and there was lots of stuff flying. And I went to my chief of staff and said, okay. Look, I have. All this stuff is happening. Tell me how you can help me manage all of this, right? Help me, what is your take on? What should I be prioritizing? What tasks can you perform? And did all that. Nothing I could disagree with. But then the interesting thing happened. And the agent proactively said. Leonard, rather than telling me the tasks you need help with, tell me what your objective is and I will figure it out for you. Right. Right. So you don't need to detail the tasks. I'm already doing that. Just tell me what your objective is.

Karen Lynch: Right.

Lenny Murphy: And that was just like, man, I just don't know I'm ready for that yet. Right. The capability, and there is nothing based on my experience of utilizing these agents. I have no reason to doubt from the evidence available to me from my experience so far that it could not do that in a kick-ass way.

Karen Lynch: Right.

Lenny Murphy: So, why wouldn't I? Right. Right. From an objective standpoint, why do I spend, why do I continue to spend time in managing processes that could do better?

Karen Lynch: Right certainly faster right than me uh overall and that could include you know financial components I'm not ready to do that yet right but boy I can see it coming pretty quickly right so that master card video again they it took a 400 budget so it showed how it broke down you know and it did max it out That was very generous of it to not max it out for our sake. And I think for our sake, the viewers of the video. We needed to see that it didn't max it out. Otherwise, we would have been freaked out because we would have said, Okay, well, if we gave you a thousand dollar budget, would you have maxed that out? You know, we needed to see in these early video demonstrations that it wouldn't max out $400 so that we could even comprehend if we gave it more than that. I think it used to be somewhere between, you know, uh. Probably got up to between, you know, just over $100 or between one and two, maybe, if I'm doing the math right in my head now. But anyway, you know, it just showed you very clearly how it was using the budget. But after the fact, you know, it did the math. It was showing us its thinking for the demonstration. But if I was just the user, it would just have gone and done all of the things and I'd see the end result. Do I want that if I'm? I'm procuring creative? Do I want to see the end results? I don't know. I don't know. Maybe. Maybe I do. Maybe I want to just give a MasterCard a budget and say, go get me some creative and have it bring me something. Do I want that? I have, that's why I say I would like people to go watch this master, people in our industry, watch this MasterCard video and then. Get back to us and say, Yeah, I was pretty freaked out too when I started to think about it because it unsettled me for a different reason. It unsettled me because I was like, maybe I would do it if I got too busy.

Lenny Murphy: Yeah, I would. I'm a results guy. So, here, just emblematic of my thinking on these things, right? I hate putting together furniture. Yeah so you know so oldest daughter moved back home and ready to get stuff and it's like I would rather spend more and buy a fully assembled piece of furniture than go through the process putting it together yeah because it's a waste of my time yep I get no value out of that process that there's no sense of artistry of oh look i've put the screw in right I don't care about any of that I want the result so um, and that's how I feel about cars.

Karen Lynch: I feel that way about cars. I hate car shopping.

Lenny Murphy: Yeah, this, yeah, same thing. Unabashed, you know, most things that I publish, you know, AI has written most of that from a language standpoint. I'm not a poet. I'm, I'm, you know, it's the result. I'm more interested in the result and what happens with the result and the impact. And that, that is what this, all of this is about. You know, uh getting to the result yeah um and it is revolutionary uh in in so many not just business but also from the data standpoint because to your point you know who else is going to get access to the that agentic record so not the brahm not not circana not any of that master card that mastercard I know that's you know linear caren's um but the other folks don't anyway it's so interesting and at the same time yeah yeah and then microsoft uh is treating agents as insider threats um yeah so uh so they're setting up agents to manage agents right who watches the watchers um yeah uh but gartner reported 71 of organizations expect to deploy genetic aid within a year while only 21 have mature governance And that's the other piece that we're talking about: Rails and governance, et cetera, et cetera. It's not just the hardest for what the agent does, it's also managing the data and the provenance and the proof of authority.

Karen Lynch: Yeah.

Lenny Murphy: All that's just all a great big deal.

Karen Lynch: Yeah. Yeah. Well, certainly, I hope Microsoft is in that 21%. I mean, not Microsoft, MasterCard is in that 21% because. We're putting all those agents in charge of budgets, they sure as heck better have governance in there because it's terrifying.

Lenny Murphy: Anyway, it is. It is.

Karen Lynch: All right. Let's just rattle off some of these. This is stuff you can Read on your own, folks. It's really just, yeah, deep reading on some research, right?

Lenny Murphy: It was all about synthetic. Right. I think all of these things from variation on theme, the human behavior bench by rehearsals from Google, Deep Mind and Kagel. How well AI predicts real people's survey answers and gives synthetic. The next interview playground, John Hopkins and Listen. Tests AI interviewers in simulation.

Karen Lynch: So a different version of synthetic.

Lenny Murphy: Uh and then this last one I thought was really useful mx8 labs yeah I guess that's that correctly um uh the methodology guide uh huge uh they just collected all types of uh the academic papers uh on all of this really encourage everybody to to look at that um these sources uh I love that there's so much research being done now on this And to be clear, the takeaway, my takeaway to it, you think is that it's fit for purpose. So no, I agree.

Karen Lynch: I think it's fit for purpose. And I think it's fit for purpose and only as good as. The goodness that goes into it. It's only as good as long as it's being updated. And it's, you know, I like the one-to-one correlation that I hear. And I like that the one-to-one is the one that it's based on is being updated and that's being validated. So fit for purpose and a validated twin.

Lenny Murphy: That's right.

Karen Lynch: It's got to be quality.

Lenny Murphy: Yeah, quality. Gigo. In, garbage out. And you need a way to make sure to track the garbage, right? I mean, we're doing supply chain, data supply chains now. So, right, right. Anyway, yeah, busy week. I'm going to miss you next Friday.

Karen Lynch: Oh, well, we will miss seeing you there next week, but we'll be updating. We'll be updating the people back home from the road. I'm sure you'll see pictures of the venue once we get it all set up.

Lenny Murphy: Can't wait.

Karen Lynch: Yeah, we'll let you know. We'll let everybody know how it goes. And if you are listening and you come, we will see you there. Matt, I you're listening. See you there. And yeah, we'll put on a great show because, you know, the attendees are who make it great and our sponsors make it great. So thanks to everybody.

Lenny Murphy: We'll see you there. Absolutely. And then we won't have any events to talk about for a little while.

Karen Lynch: So I mean, APAC will be here, you know, before that's true.

Lenny Murphy: That's true. Yeah, it never stops.

Karen Lynch: All right. Yeah.

Lenny Murphy: Everybody have a wonderful weekend. Thank you, Karen. Thank you, audience. Hope this was useful to everybody, and we'll talk soon.

Karen Lynch: All right, bye. Bye.

Links from the episode:

Salesforce buys Braze for $5.66 billion 

Qualitate Raises $25 Million Series A Led by Insight Partners 

Day 53 Update on Acceptance Levels 

Alfred Wahlforss - Listen Labs launched Listen Twins 

Highlight launched Pulse 

Circana Puts Third-Party Marketplace Sales on the Map: $200 Billion of Previously Hidden Demand, Now Measured Weekly 

Angelique Taylor 

While Amazon blocks AI agents, Walmart rolls out the welcome mat 

What happens when AI agents start paying each other? Watch Mastercard Agent Pay for Machines in action 

Microsoft says AI agents are like insider threats as ‘frontier models can’t be trusted’ 

Gartner Says AI Risk is Rising Faster Than AI Value 

The Human Edge in AI: GLG Frontier Powers AI Labs with Critical Expert Judgment 

Rehearsals Launches Human Behavior Bench With Google DeepMind and Kaggle 

InterviewPlayground: A Validated Simulation Environment for Evaluating AI Interviewers 

MX8 Labs - Can I trust this number? Synthetic data and statistical inference 

Comments

Comments are moderated to ensure respect towards the author and to prevent spam or self-promotion. Your comment may be edited, rejected, or approved based on these criteria. By commenting, you accept these terms and take responsibility for your contributions.

Karen Lynch

Karen Lynch

Editor-in-Chief at Greenbook

394 articles

author bio

Leonard Murphy

Leonard Murphy

Chief Advisor for Insights and Development at Greenbook

797 articles

author bio

Disclaimer

The views, opinions, data, and methodologies expressed above are those of the contributor(s) and do not necessarily reflect or represent the official policies, positions, or beliefs of Greenbook.

More from Karen Lynch

 The AI Shopping Gap Nobody Is Talking About
The Exchange

The AI Shopping Gap Nobody Is Talking About

Why do consumers trust AI for shopping advice but not purchases? Explore AI shopping agents, retail ...

IIEX West Recap: Key Takeaways with Margaret Mariani
Available on YouTube, Spotify, Apple Podcasts, Amazon Music

The Greenbook Podcast

IIEX West Recap: Key Takeaways with Margaret Mariani

Play episode

Karen Lynch and Margaret Mariani recap IIEX West, sharing event highlights, industry trends, research innovations, and key insights.

Your Next Customer Might Be a Bot
The Exchange

Your Next Customer Might Be a Bot

Would you trust AI to test your ads or shop for you? The Exchange explores synthetic research, shopp...

Datacy and Enlightn Win at the IIEX West 2026 Insight Innovation Competition
Greenbook Company News

Datacy and Enlightn Win at the IIEX West 2026 Insight Innovation Competition

Datacy won both the Experts' Choice and Audience Favorite Awards. Enlightn won the Industry Impact Award. All five finalists took on problems practiti...

Sign Up for
Updates

Get content that matters, written by top insights industry experts, delivered right to your inbox.

67k+ subscribers