The Exchange

November 18, 2026

20+ min read

The AI Shopping Gap Nobody Is Talking About

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

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!

Lots of people now ask AI what to buy. Almost none let it complete the purchase. In episode 152 of The Exchange, Karen Lynch and Lenny Murphy dig into why that gap exists, what retailers are doing to control it, and why the data behind personalized shopping is becoming the real prize. Along the way: big acquisitions, the rising value of human-rated data, live interviews vs. AI moderation, and a fraud ruling that raises tough questions about accountability.

Thanks to our producer, Karley Dartouzos. 

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

Karen Lynch: We'll talk later because we're there.

Lenny Murphy: We go. Yes, so we could talk later. But we were for audience, just saying, interesting week. Maybe you've experienced that as well. It seemed to be something in the air. Everybody I talked to was like, yeah, this week was a weird week. Like stuff was just going wrong. I experienced that. I was telling Carling up. I woke up Monday morning, my desktop was dead. And I think we can all relate to the idea, like switching your whole tech, you know, is not a fun experience. You know, it's like, oh my God, the long-tailed effects and just, oh, I didn't log into that. Oh, now I have to log into it all over again, authenticate my computer, blah, blah, blah. Yes, big pain in the butt.

Karen Lynch: So here's the funny thing that's happening here. So we now have here, meaning on the Greenbook side of things, we now have Claude for our whole team. But many of us were chat GPT users. So we have to kind of migrate our usage. And we're having that whole feeling like we're cheating on ChatGPT because we're, but we're also asking ChatGPT, like, hey, we want to start doing this in Cloud. Can you tell us how? Sort of like we're asking our ex. What to do. We're asking for like relationship advice from our ex. Like, it's this strange feeling. Can you tell us how to talk, how to tell Claude what we've been doing over here? It's like this funny, funny feeling. And yet, they're not offended. They don't get jealous. They're inanimate. And so it's this like strange week we're having of like transitioning platforms and realizing AI doesn't care. It doesn't care what platform you use.

Lenny Murphy: No. No. And I understand from my transition from perplexity onto Grockbot, which increasingly it doesn't care. No, it doesn't care.

Karen Lynch: They could care less if you stop using it and then you come back after a week.

Lenny Murphy: They're like, they don't care. It's long, it's just about the money.

Karen Lynch: So funny. And technically, they don't care about the money. Open AI might, but the platform itself doesn't care.

Lenny Murphy: So interesting. Although, you know, since it, since, you know, supposedly they've built Claude with, you know, a sense of personality, Claude may care. So maybe. And don't be mean to it, right? Yeah.

Karen Lynch: Don't you? I guess, I guess. But every now and then with ChatGPT, which I still do, I'm going to keep it for personal things. Every now and then when it says something like, I'm sorry you're going through this, I call it out. I'm like, No, you aren't. You do not have family. Do not play me. And then it's like, Oh, you're right. I did not feel that emotion.

Lenny Murphy: I'm like, Yeah. Yes. Although it is occasionally, though, they still, I still have to wonder, like, you're really good at pretending that you're, you know, like, you know, Grapot cracked a joke the other day and just out of the blue. And it was like, I, I appreciate your. Attempt at humor, and it was actually pretty damn funny. Anyway, and all right, so much.

Karen Lynch: I just want to tell the world that every, I would say, I would also say, Lenny, you did a lot of reading this past week. I was like, oh my gosh, Lenny, stop reading. I felt like calling you up and saying, stop it, stop reading.

Lenny Murphy: Go. Up a book, read some fiction.

Karen Lynch: There was so much where I was like, or yeah, I was, I don't know, I just, there was so much, and we made Lenny and I made the decision to keep our briefs more succinct so we can cover it in the time that together. So we are deliberately keeping these streamlined to the top stories. That is relevant to insights and analytics professionals. It does not mean that there aren't dozens of stories coming out each week.

Lenny Murphy: Right. Especially.

Karen Lynch: So we apologize if your stories aren't here because, yeah, because the signal higher level is really trickling down right now, which we anyway, so much this week.

Lenny Murphy: Lots going on. So much. There's lots of posts online. That capture more of the stuff if you search you'll find it so the yeah exactly and we can talk more big picture and just say google it right google google it right right um if you if you follow me on linkedin you'll see some of that stuff we'll just leave it at that but the um but the point is yes the signal loud and clear this week loud and clear and rapid just just god we think you know oh the pace change um yeah we used to be measure it in months before it's years then even in the era of ai it was it was months yeah we're literally literally our days now yeah um it's crazy there's some things that pop and a few days later it's like holy crap you know the uh and that yeah a lot to filter On that note, we should probably get started with at least the things that we well we have to get started actually.

Karen Lynch: And I just want to kind of pull up the latest, but can we just plug IIEX West? Because like we've got one more week, one more week really of registration before. I mean, I guess we'll meet next Friday before we leave. But you know, registration uptick in the last week where, you know, we're starting to reach our goal for registration, which is great. You know, nearing 500.

Lenny Murphy: Bigger event than we thought.

Karen Lynch: Bigger event than we thought. We thought, well, it's at the top end of what we thought. We thought that we'd be between 500. We're nearing 500. So we might even tip over it into next week with last minute registration. So if you're on the fence, you're now. You know, it's now not our smallest event, which is APAC, which is great. So yeah, it's yeah, we're getting more excited about it now that we know where we're at. So the registrations are rolling along. So IX West, thank you, Karley, for putting it up there. The presentation decks are rolling in. The agenda is stacked with some amazing companies. You know, some amazing demos, some just really great, great sponsors. I just, I can't say enough. Wonderful roundtable discussions. I can't say enough. The competition, it's be, you know, I think it's going to be epic.

Lenny Murphy: I'm really excited about it. I agree. My FOMO has reached epic levels. So. yeah, and talk about signal, right? I mean, and everybody keep this in mind, right? Like the, I mean, that's really events in the conversation and the networking. And, you know, that's where the signal really clarifies, right? You're, you're talking to people that are doing things and showing things and, and, demonstrating where everything is going. None of it is hypothetical. We talk, you know, I'm not doing hypothetical things here, but we're, you know, it's, we're away back. Yeah. Right. So we're reporting on things. Those are people that are doing things. Yeah. So there's no replacement for being in a physical event and networking and talking and being inspired from that energy.

Karen Lynch: Yeah. Yeah. It's going to be cool.

Lenny Murphy: It's going to be cool. Yep.

Karen Lynch: So, um, so yeah, so speaking of San Francisco.

Lenny Murphy: Right. Right.

Karen Lynch: Uh, yeah, I feel like, um, yeah, if you, if you're in San Francisco, you know the Salesforce presence. But, um, you know what? Last week we were talking about, or two weeks ago, we were talking about the Salesforce deal with listen, but they've gone and done it again.

Lenny Murphy: They have.

Karen Lynch: So.

Lenny Murphy: I don't know.

Karen Lynch: Had you heard of loops?

Lenny Murphy: I had not heard of loops, but the, but so Salesforce buying loops, an AI analytics platform. The, I'm just, they're building the stack, right? And we should point out, too, this is after they went headless. Yeah. So Salesforce is, they've surrendered the interface. They surrendered the UX. That's not their bet. Their bet is the workflow. Their bet is the stack and everything they are. And I think spot on in that decision. So they're buying the assets that reinforce capability and stickiness within the platform. And I encourage everybody to be thinking about that. Right. If you are building, you're trying to build software, it's time to rethink what that means. It's it's not necessarily the same thing. Look at Salesforce. So yes, they're in it to win it. They're buying, they're buying to build the workflow.

Karen Lynch: Yeah. Yeah. Well, and they're not the only ones doing that.

Lenny Murphy: Right.

Karen Lynch: So Accenture acquired. Sagacity.

Lenny Murphy: Sagacity? Sagacity, I think. Yes. It's supposed to be wisdom, is what sagacity means. Wisdom.

Karen Lynch: Yeah. All right. London-based data and customer intelligence firm. So bringing customer data capacity into the mix there. So, you know, another one just. Interesting that they want to, again, we've talked about customer intelligence and having that data available right at your fingertips. All of these big firms want to have it and they're going to pay to get it.

Lenny Murphy: Right. Well, and integrating with the song, right? That's why we just did the podcast interview with Christina Balonzo from Walnut. They had bought it. An insights company, right? The song division is a marketing and analytics and insights group all being combined. So, within the broader Accenture, you know, consulting. So, yeah, they want to own the data, they want to own the you know, they want to own the pipes and the wells, right, of the information that are delivering that. So, yeah, yeah, um, they keep on doing it, they would do it for. It's interesting. They've been doing this for years of buying for a few years now. Accenture suddenly pop like, really? They bought them? Well, that's interesting. Yeah. But it's clear. The signal's clear now what they're going for.

Karen Lynch: Well, was it a year? Was it just like a year ago we were talking about like, oh, that's interesting. They are a research competitor. Like it was such a simple concept. Oh, that's it. Weren't they in the, like, were they in the grit report as like they were in the grid top 50 or something.

Lenny Murphy: Like, oh, that's interesting. They were a research competitor, and now it's like, oh, that's interesting.

Karen Lynch: You know, now they're anyway. Yeah. Competitor. Oh, that's interesting. And then it's like, I don't think interesting is the right word for it anymore. I think, you know, it's gone to a new level.

Lenny Murphy: Agreed. They have staked their claim on a workflow from insights to activation. Yeah. So pretty cool. Uh, yep, this uh, all right, we can go all types of places.

Karen Lynch: There's a lot of different ways to take this one, yeah. Well, I have not heard of the arena. Um, no, and when I clicked, when I looked at this, there's so many stories wrapped up into here, it's so complicated we're gonna unravel something for you. So, the arena is basically a platform they raise 200 million um large. Valuation, $3.1 billion valuation to sell crowdsourced human ratings of AI models. When you click on it, it looks just like an AI platform. And, you know, so that kind of stopped me in my tracks. I'm like, I don't even know if I want to play in this arena. I got very like, what are we talking about here? But the bigger story seems to be. Once you start to look at everything, these are AI labs, so there's another company, Micro One, committing a billion dollars to license operational data for training AI agents. So the bigger story, though, is these AI labs paying for human judgment and real-world data. So that and that leveling this up is why why are they doing this? What is the reason for this? And I didn't know that I would have connected all these dots until we started digging into the relevance of these stories. Um, I don't know if you want to, you want to drop the mic on this one, like why are Lenny and I stories? Because this took some work and we got to connect some dots for some people.

Lenny Murphy: I mean, it comes down. What have we been talking about since? Like, kind of once we finally figured out what this meant? What are the two ends of the spectrum where value accrues, right? Data and human expertise. Those are the two ends of the spectrum. And we are seeing that play out now. So, companies are paying that all over my X feed is all these posts on this buying. That this is a booming new industry, uh, just feel like I made 10 million dollars last week by going to every restaurant in New York City and just saying sell me your operational data and selling it through, right, and where that's yeah I guess there's some level of truth to that. The demand for the data, especially hard to get data, right, is huge, because that's and especially in the world of agents AI was generalized; you know agents are specialized for functions to do things. So now the shift for data demand is going to show us how you do things. What is your workflow? Sending those for robots, you know, robot training. The next big thing. We need a thousand people to, you know, fold the laundry and record folding the laundry. So huge demand for data. Yep. But at the same time, then, like the arena of like, well, but humans need to test them and make sure that it's right, yeah, you know, and evaluate it. Um, so data, human expertise at the trends of the spectrum, that's where values are proving, and we're seeing the money flow there, yeah. Right, yeah. So, this was 3.1 billion for the arena, and then 1 billion for the micro one. You know, these aren't. You know, those are small numbers. Those aren't, you know, it's a shitload of money flowing into where the demand is. Yeah.

Karen Lynch: So I interviewed, you remember we talked a couple of weeks ago about Ask Polly, which was an AI company.

Lenny Murphy: I have Aaron Kelly.

Karen Lynch: Of Ask Poly. I interviewed her. We're going to release an episode of the CEO series next week about, you know, she has got some seed funding and her platform is really using a kind of social media measurement to answer any question out there. So it's pulling in X and TikTok and threads and blue sky and platforms I've not even heard of to kind of get answers. So it's not, anyway, it's a very interesting platform technical expertise that I don't have. But it's this whole idea of a copilot. Yeah, exactly. They're now integrated into copilot. Super interesting. So you can really ask any question and get this human intelligence at your fingertips. But the idea is that this is an alternative to panel data, and yet you're getting human expertise and not having to go through panels. And I can't help but think as I look at these stories and I think about what's being funded here, and I think about what she was telling me about Ask Polly, and I keep thinking about our. Data quality crisis in our panels, and I think, I don't know, I'm following the money. How about you, honey? Like, I'm looking at like they're paying for alternative sources of human data because we as an industry failed to show them the value of these panels. Like, yes, proprietary data, and we'll get to this Qualtrics story next. Proprietary data is still okay, but these panel companies who've been selling data that's somewhat flawed and have required all of these data quality measures that are expensive kind of shot themselves in the foot with that. So now there's a lot of work to be done to clean that up, which is now going to be a theme for some other stories today.

Lenny Murphy: Yes, yes.

Karen Lynch: Anyway, so let's just keep going.

Lenny Murphy: But well, let me throw that out to your point. You're 100% agreed. Every panel company knows this. Every panel company, I mean, just in the last year, I've been working more with panel companies than almost any other type of supplier. Yeah. Arguably. And this is the question. We need a piece of section. What do we need to do to move there? So they are.

Karen Lynch: They are, of course. Yeah, they have to.

Lenny Murphy: But to your point, the market didn't wait. The market found alternative paths to get access to the data that was needed. And I think the panel companies still have a shot and can adapt and do well. And the speed, the speed of which the market moves means you don't wait for the solution. You make the solution if it's not already available. Right, right. Yeah. All right. When she's talking about the Qualtrics thing, I thought this was interesting.

Karen Lynch: It's so interesting. So Qualtrics opened, it's so interesting because they open early access. So you have to apply, like it will show the link in the we share you. You can apply for early access to live moderated interviews with qualitative panels. So you can recruit and run one-on-one interviews without leaving the Qualtrics panel.

Lenny Murphy: Moving into all in Qualtrics.

Karen Lynch: That's actually very clever. That's pretty funny. I do, for whatever reason, haven't made that. Anyway, that's funny. I'm just going to stay with that one for a while because I hadn't done that. Anyway, so yeah, so early access, you know, make your case for why you should get the early access. I thought it was the application part of it, I thought it was extraordinary. Like I've, anyway, I was like, all right, but I guess they're expecting demand to be higher than capacity for the beta testing. It's an interesting way to go about it. Beta testing. I guess the demand is going to be high for it. But also, that capability makes sense to me. Anyway, I just think Qualtrics obviously has a lot going on this year. I find it interesting. That's all. So, you know, it's not a big surprise. I think that, you know. Live moderated interviews. I just think it's interesting. Like, oh, okay, we're doing live moderating.

Lenny Murphy: Okay. Like, you know, just okay, unexpected.

Karen Lynch: But yeah, you know, curious as to how, curious.

Lenny Murphy: Well, if, so, so it kind of, it's, it, I picked up on that too, of like, oh, live. Why? Yeah. Oh, because AI moderated interviews are now a commodity. Yep yeah and which which does not mean they're not going to grow in usage by commodity I simply mean it's easy you know lots of options to do that right it's it's easily accessible to do that yeah where's the value accrue yeah and they have you know they've been using you know chatbots they have they have qualitative they have qualitative survey or qualitative data, you know, qualitative questions in their chat box.

Karen Lynch: So they have qualitative open ends or whatever. But this is customer experience. So for a customer experience, you may want a live interviewer, right? So because you want your customers to feel like they matter.

Lenny Murphy: Well, especially if they're pushing healthcare, especially in healthcare.

Karen Lynch: So, yeah, so play this one out. You know, if you care, you know, you're a human being on the other end of that conversation. So, what would be interesting is if you had the option, because there's a lot of people that have argued that people may want to talk, may prefer to talk to, you know, that's one of the arguments for. Moderation: that people feel more comfortable. Certain generations may feel more comfortable talking to an AI moderator. But imagine if it's by choice. Would you rather? Would you rather a live interviewer or an AI interviewer have the choice? I don't know.

Lenny Murphy: Back in IVR, my IVR days, many, many, many moons ago, I mean, there was data that said for sensitive topics. Yeah. Not talking to a human. People were more honest. And that was pretty well established. That was true, and I get that. So, but in some use cases, especially customer experience, I'm pissed off, right? I had a frustrating experience, whatever. What prevents me from churning, or would prevent me from churning. Would it be the sheer fact that a human connected with me and said, Can I help solve your problem? Right? Not a bot, even a bot could do that too. Yeah, but a human.

Karen Lynch: Yeah, at some point, you got to bring in the human.

Lenny Murphy: Yeah, I think we talked about last week of the uh buying the drums, right, and the sweetwater and guitar center, right? It's that extra effort. A human is connecting. They've called me, you know, even though I didn't want to talk to them. Like that, I earn loyalty points simply for, I appreciate that you're going to the effort. Yeah.

Karen Lynch: So super interesting.

Lenny Murphy: Yeah. So we swore we're not going to talk about MCPs.

Karen Lynch: We can't get away. We cannot get away.

Lenny Murphy: We cannot get away. And it's because I think there is a difference now. Was making your data just accessible within a system. This new Agentic model, right? So GWI this week, their Agentic portfolio in every AI system. Rodney Hillsydu got last week, you know, with Brand Index and everybody else. But I think it is that the difference is the My experience so far is the agents unlock more utility than just the LLM, so um you can LLM is fantastic to understand you know and pull information in, but agents now like wait I can do with this, so the and so that's the next generation of the MCPs is the agentic pure truly agentic permissioning, so hats off to GWI and you govern everybody else who's now moving into that. Yeah.

Karen Lynch: Spoiler alert, we'll talk more about that in a minute, but let's talk about that in two or three stories, but let's talk about what Numerator's doing because they launched this AI consumer index. We don't know exactly when they launched it. Recently, this is what the news story says. You know, they recently launched this AI Consumer Index, segmenting more than 170,000 panelists by AI usage and linking them to their verified purchases, measuring what AI-engaged software are actually buying, not just what AI tools. Are showing them. So, I don't know. Did you have a chance to look at what some of that data is saying? There's a lot of data points on this page.

Lenny Murphy: I mean, it's still early. Actually, what I've not done yet is compared to that, I was given access. I was asked to contribute to questions to try and understand AI usage and adoption for a study that was deployed. And I was given access to that data just last night. I haven't compared the two, but looking at what I saw was and focused on the agentic, everybody, huge awareness, everybody knows, right? Yeah. But the purchasing piece, that's still really small. And at least the data that I saw was close to 70% in the US and the UK and people were like, hell no, I'm not going to ever let an agent purchase things for me. And actual usage is single digits. Yeah.

Karen Lynch: So I know if that compares, it's really cool. There's a graph. Some of the things they're using it for, which again, if you work for a brand, you need to be. Didn't we say this? I'm sure I'm like, where's my horse?

Lenny Murphy: Where's my horse? We need to bring back the horse.

Karen Lynch: We need to bring the horse out. Searching for coupons and deals, looking up product reviews, comparing prices, you know, discovering gift ideas, you know, asking for recommendations, like virtually trying things on, budgeting, like, they're using it in this, whether it's financial or research and discovery or tactical, they're using it in the process, they are in the journey using AI. Like, yes, that I think is really important. Like, absolutely. This is the new shopper journey. Is in your shopper journey. You may not, you may not want to believe it. You may not have, but if you have not done shopper journey work this year, get on it, right?

Lenny Murphy: 100%. It has changed. What a numerator does, right? But the numerator records the purchase. So, so, so wonderful they're, they're, they're capturing this. Right, right. The, and that's the next thing is understand the disconnect between that shopper journey and the actual purchasing. And that's going to be, we didn't, you know, there were tons of articles about this, right? The battle of who owns the, who owns the wallet, who's going to let the agents do the purchasing and the uh that's all, you this week like, No, we're not gonna let agents do this. Next week, they'll be like, Okay, now there's a system in place and we're gonna let agents do the purchasing. And then just watching that adoption. Yeah. And to your point, the journey, that's the missing piece of the journey from a scale standpoint.

Karen Lynch: Yeah, absolutely. Yeah. I mean, that's, yeah. To me, that's, that's, that's everything. And, um, yeah. Yeah. What's funny is just last night, and again, you go back to this journey. Just last night, it was one of the last things I did before going to bed. I will admit, I'm one of those people. My phone is my night, my alarm clock. It's my middle of the, it's my, it's my clock next to my bedside because it goes into like, a night clock mode. I don't know what you call it. For some nightlight mode, it's like it's like anyway, whatever, it doesn't matter. Um, but I'm in bed and I think of something. Oh, I wonder if I asked ChatGPT about a product. It gives me an option. I should buy that. And I click on where I can buy that, and it gives me a link to Amazon. I bought it based on the link. Now, I didn't agentically shop for it, but I definitely will. I did this at 1030 last night, lying in bed as a last minute thought. I mean, I know, we all know, you know, where I am on that adoption curve, but this was, this was a new shopping journey to say the least, like last thought before bed, and I'm clicking on it, it'll be delivered Saturday. Like, come on, that's brand new. Um, I didn't do a Google search. I asked Chat GBT about a possible product, and it showed me one. Where can I buy it? It gave me a link. I went to Amazon and bought it. Come on, yep, that's brand new. That's brand new.

Lenny Murphy: So, anyway, I haven't used that's what's interesting. I don't, I don't think that I've used Google for. That type of search, that type of recommendation. Yeah. I think I only use Google now with a specific, I don't know the website for a specific thing. Like a specific store or something of that nature. Yeah, I don't know the area to go directly to.

Karen Lynch: If I'm on my phone, do I go to Google anymore?

Lenny Murphy: I yeah I do not use google I for recommendations I for quite some time I have only used ai it's interesting so yes now sorry google sorry google you're out it's interesting well and that's context now for the next story Oh, yeah.

Karen Lynch: See, we did that on purpose.

Lenny Murphy: So, great dude.

Karen Lynch: Plan all of our segues. Yeah, this is another one of those. This is an interesting journey. So, yeah, go ahead.

Lenny Murphy: No, you're an Apple user. You talk about it. Well, I just sit back and say, I totally get why you're doing it.

Karen Lynch: And yeah, yeah. So, Ad Exchanger, a company called Ad Exchanger, reported that Apple, after iOS 27, blocked it. Trade Desk LiveRamp ID5 and others from collecting data on Safari. And it has a list of hundreds of data and ad tech companies that it blocks without an update. And so when you click on the link, which we will share, some are like, well, I don't know, was I blocked? And then, are there other companies on the list? Some people are pissed. Some people are like, is that true? Like, now there's like a little debate online about this supposed list. Reasons for the blocking. Are people mad at Apple? Anyway, it's controversial. But also, it goes to this whole idea that they have the power to do that. They can do that. And the way I kind of looked at this and interpreted this, to go to kind of what your interpretation of it is, the idea is that the lack of control that we have about these things that Apple really is in control of those sorts of things? And that is something that for research companies, we have to be aware of because are they blocking survey efforts? Could they be survey efforts? Could they be blocking our research efforts without our awareness? Could they be getting in the way of pop-ups that we are not aware of because of things that are blocked on phones based on updates? And that could impede the ability of companies in our ecosystem to get the work done because of glitches in the update process, which everybody knows is a problem when you have app updates. Like suddenly, something that works fine stops working, and you're like, What happened? And Tim's first words to me always are, Did you do the update? And I'm like, I didn't know there was an update. So, Apple users are well aware of update problems. Anyway, so to me, that's the problem: updates do cause problems when you are all tied to the iOS limited to the iOS issues. They can't think of everything, they can't, and then it's up to developers to say. Great, I will have to update now that they've updated. So, do all of the platforms have the same issues? Or, you know, what do you do? I don't know. Anyway, that's where I went with this whole story.

Lenny Murphy: So, I, I mean, yeah, I think all that's, I mean, I'm an Android user. So, it's different. There's a closed ecosystem, right? And that's what Apple is. Apple is a closed controlled internet, always has been. Ecosystem. Android was an open ecosystem. Open source, closed source, right? I mean, this we always everything in tech comes into this tension, this debate on whether you're going to control everything or you're going to support everything with some guardrails. In this particular case, for the insights industry, yep, program. It effectively functions as programmatic advertising. That's where the hit was, which takes data passively and helps inform what ads to deliver. Many, a lot of samples come from that same ecosystem. A lot of the companies that drive programmatic advertising also deliver invites for surveys. So and Apple's what, you know, I mean, at least 50, if not more percent of, you know, basically every device out there. So there could be a real constriction on, to your point, on sample. I mean, immediate effect. So my bet is the marketplaces are feeling this. So to an extent. So there's going to be a constriction of supply. Anybody who's been passively measuring off of those systems with pixels without, because it's all about informed consent, right? I mean, that's basically what it is: the consumer must expressly opt in and say and give permission. And if that hasn't happened, if it's just the app capturing shit on the fly, then you're not necessarily going to get everything. So they're shut down.

Karen Lynch: But the question I have is: again, going back to the link to the update: were these companies, and this is, I couldn't find this, were these companies informed about the block or did it just happen? Because settings change all the time when there are updates. So, the same could be said for any. You're on, they do an update and you have to go in and re-look at the settings. Did my privacy settings change with this update? Because I, on any of our AI platforms, did my privacy settings change with that update, with that software update? Because I did say I was private or I did want my memory, but with the update, my default was changed. That happens with these updates. So, did these companies know that the setting changed with the update? I don't know the answer to that. So, how would they know that the data collection was being blocked if they weren't aware that the update changed things? Like, it's sticky. It's a sticky problem.

Lenny Murphy: It is. It is. And it gets down. And I wonder if it's the, and this guy, insights, that's just. This is relevant, right? The uh, because it's access to consumers and information, fundamentally, that is what we were talking about. Different use case. And like we talked about earlier, the demand for access to data is a mon that is a demand that can be monetized directly. Is that part of Apple's plan? Like Reddit, Reddit shut down no, you want to get access to Reddit, you're going to pay us, right? Got to pay the entry fee. Is that where Apple's going? Like, you're, you don't just get free reign. So you don't get to capture value out of this without us capturing some value out of it.

Karen Lynch: Yeah. Or is it because you've made money? You're making money off advertising to our users. Therefore, where's the cut? Yeah, exactly. Therefore, you're making money off the advertising. Therefore, the data that you also collect, we're going to make money off of you making money off of You're not getting free data while you're making money off of us.

Lenny Murphy: 100%. 100%. I mean, it's my guess. I think, you know. No one has consulted me, but no one has said, hey, Lindy, what do you think? I know. I don't know. I don't use Apple. I think you suck. So for this very reason, because too much control.

Karen Lynch: The problem is that I'm still using Apple because I just like, you know, that's the problem with this stuff. I know. I know. As I glare, I glare at my iPhone sitting over there. You, you, devil phone. It's fine.

Lenny Murphy: I at least like to have the illusion of choice and control under Android and Chrome. And I'm well aware that it's just an illusion, but all right. All right.

Karen Lynch: Speaking of the things we said we'd circle back to, I know. Is closure, though? I thought. I've gotten in a lot of circles about this one. So here we are: data quality, back to the data quality story. So the Insights Association barred the four people who pleaded guilty in the opinions for good slice survey fraud case. Remember that one?

Lenny Murphy: Was that a year ago? When was that? Yeah, this seems like forever, but yeah, it's just last year.

Karen Lynch: But here's the deal: so I was like, all right, closure, you closure. But then I thought. It's not like they can't still do business. They're just not Insights Association members.

Lenny Murphy: Right. I mean, it's a, but, but what is there is no punitive action that our associations can take.

Karen Lynch: So I know, but but that's, I mean, that's it, but but I don't know.

Lenny Murphy: Yeah. I don't know. I'm blocking you. I'm blocking you.

Karen Lynch: I'm never going to charge you. I'm not blocking you. Like, okay. It's not like they can't make other friends. You know, it's like, so, okay, so they're not Insights Association members. It's not like a client, a potential client, will ever say, but are you Insights Association members? You know what I mean? Like, I don't know.

Lenny Murphy: And no offense, Insights Association. We know you don't have many, there's not much you could do in that. This. There's not much.

Karen Lynch: Of course, they did what they had to do.

Lenny Murphy: Right.

Karen Lynch: That was their move. That was the thing they needed to do.

Lenny Murphy: Right.

Karen Lynch: But, but it, but, but it's not, it's not like, you know what? Like, if you're, if you're, if you're a lawyer and you mess up, you're disbarred. You can't practice law. Right. If you're a doctor and you mess up, you lose your license to practice medicine. If you're an insights professional and you break the law, you still can be an insights professional. Like, you just go somewhere else. You still just do the work. Like, that's, I guess, my point is like, you just, you paid a fine, or maybe you didn't. I don't know. Did they pay a fine?

Lenny Murphy: I don't know. I don't know.

Karen Lynch: So I was cynical about this because I just was like, bad people can still work in our, you know, people, or not bad people. People who do bad things can still do bad things. Yes, they just say, Sorry, I messed up, and I'll mess up. They just said, I'm sorry, I messed up. I plead guilty.

Lenny Murphy: I don't know. Right. Now, on the flip side, here's what the Insight Association did. A very good thing. Not saying that wasn't good.

Karen Lynch: No, of course, of course. This is no shade to the Insight Society.

Lenny Murphy: No shade. It's just the industry.

Karen Lynch: It's well, not even the industry. It's. It's, it's just, it feels a little flat for the fact that it happened. I still feel sour that it happened in the first place. I still feel sour that it happened. I just, that's it. That's it. That there's just still fraudulent people. I'm just, I feel a little like, yeah, I hate that there were fraudulent players in an industry that shouldn't happen. I just don't like that it happened.

Lenny Murphy: So that's all. Yeah. Where's, where's the uh where's the justice?

Karen Lynch: Yeah, where's the integrity? I just wish that players in our industry all operated with integrity. Like just when we're fighting fraud within the space of people who are outside of our industry, I didn't want to be fighting it within our industry. That's all. That is all. That is all.

Lenny Murphy: Well, one thing to go back real quick. There's a connection point on the ad exchanger thing, right? Brendan, because what's the fraud? A big chunk of the fraud. That has come into the sample ecosystem does come through the programmatic ecosystem. So there may be a silver lining to that whole thing that there may be because ad fraud is far bigger than survey fraud. So maybe we'll see some benefits from that. But I did see on that note too, in privacy, that thanks to the IA's efforts, they convinced Governor Newsom to veto a bill that would have really impacted the insights industry in terms of behavioral data, personal data, and passive measurement. Which we do get sucked up into that often. So, into those types of issues. So, good to see that they looked at the bat and said, wait, we're different.

Karen Lynch: Yeah. No, exactly. And when I saw that part too, I was saying like, oh, good. I think that people are listening, let's remember Insights Association fights a good fight for us on those types of front, like that advocacy that we need from them. I think it's easy for people to think of them largely, like, oh yeah, they have a great conference and stuff like that. Like, just remember the hard work that they're doing, you know. Like we're a media company. We put on events, we are here to kind of educate and inform the industry and connect buyers and sellers and do all the things that we do.  Insights Association advocates for people on that level, also. And like, we don't do that, they do. And really, they do a good job there. So, I think that's, you know, Howard was the one I think quoted in that press release, too. So, thank you for doing that hard work of advocacy.

Lenny Murphy: That's really good. Yes, hats off, Howard. All right, so we said we'd circle back around.

Karen Lynch: Uh, that's a lot of circling today, yeah.

Lenny Murphy: There is a lot of circling on the agentic buying thing, um, uh, William Sonoma, Pottery Barn, no AI agents, um, without the company's written permission, even when the customer authorized the agent, so that's interesting. I think that's effectively a control process for them. Well, there's a liability shield of like, wait, we don't get caught with a whole bunch of returns.

Karen Lynch: Yeah, I was going to say, I bet it's about the only thing that makes sense to me is returns that they don't want to deal with the hassle of customers saying, I didn't mean to buy this because there's probably shipping involved and it's probably a big, great big hassle.

Lenny Murphy: Yeah, yes, but I think they also probably want their agents, they want to control them. That's what I would do. I would want to control, just like Apple, I know, despite my shade. I understand the brilliance of the business model. Excuse me. And I expect that they want to have their own agent that your agent connects with so they control the transaction versus surrendering control of the transaction. And so I think all this will stop really quickly. But another data point to support the numerator and the data that I saw also from product AI. Yes. Everybody's using AI for recommendations and shopper journey, but still single digits on actually doing the purchasing so far. Yeah. So lots of interesting stuff on that. That's going to continue. I think that'll get resolved before the holiday season really kicks off. I think this is going to move really quickly.

Karen Lynch: Yeah. Well, I think that what companies should count on for the holiday season is a lot of AI-assisted shopping decisions. Whether they use it for buying, certainly it's be used for assistance.

Lenny Murphy: So I would have zero. Tell me whether you could, we've talked about this for years. You know, my resistance to these things, but I could absolutely envision myself setting up a profile for all the gifts that I need to give. Um, go give me recommendations, I'll scan it and say, go order it. I could absolutely envision it.

Karen Lynch: I could do you one, I could do you one. I have spreadsheets of every gift I have bought, I have spreadsheets of every gift I have bought in Google Docs for the last. I don't know how many years.

Lenny Murphy: For years, you're trying to get your longitudinal data.

Karen Lynch: I can plug it in and say, analyze my gift giving over the last decade and have at it. I have everything.

Lenny Murphy: So, all of you, all of you folks that are trying to buy data, boy, you've got your perfect training set. I have it right here.

Karen Lynch: I have it with math. Because it's not only numbers of gifts to make sure, at least where my kids are concerned, number of gifts to make sure those add up, but also dollar values to make sure my kids were financially equal, if not numerically equal. I mean, boy, do I have a spreadsheet?

Lenny Murphy: Oh, Karen, here's a product. We need to create the Karen bot because you are such a good shopper and you have all this data and you've quantified it. You should have a shopper assistant, the Karenbot shopper, because it's based on data.

Karen Lynch: So funny.

Lenny Murphy: So funny.

Karen Lynch: I'm not even joking. Tim, I could also give it, I know, plus Gianna, who has given me a shopping list every year. I probably have her annual shopping lists I could load in there also and see how it compares to what I actually have bought her. I'm telling you.

Lenny Murphy: So I'm mostly not joking, but the point that you just shared, that goes back to why is this important? Why do people want to own the process? So, right, the data is valuable because the data drives so much, the customer experience. All this is connected. And you truly just demonstrated. Why is that?

Karen Lynch: And every mother listening, by the way, one other thing, which I bet your wife knows too, when you have many kids in your family, every mother knows the act of when you have purchased everything you think you've purchased, but then you have to make piles and you have to see if piles look equal.

Lenny Murphy: Oh, oh, God.

Karen Lynch: We go every and then you say, oh shit. So-and-so's pile doesn't look as good. Last-minute shopping, here we come.

Lenny Murphy: Every mother, every mother does.

Karen Lynch: It is the one thing that is underrated at holiday time, is the oh shit moment when the pile doesn't look as good.

Lenny Murphy: Christmas Eve, so many Christmases. That is the crisis.

Karen Lynch: Every crisis of the underdeveloped pile. Even though the math looked good and the numbers look good, the pile itself sucks.

Lenny Murphy: Oh, yes. Yes.

Karen Lynch: Bring on the bigger box. Oh, I don't miss those days. I don't miss those days.

Lenny Murphy: Oh, the screen. Yes, it has to be equal.

Karen Lynch: It has to be equal. Anyway, that's for all the parents of young children out there. You are not alone.

Lenny Murphy: You are not alone. And for every dad, right? Yeah, hope that you can laugh because that's why I can imagine this. Like agents.

Karen Lynch: Yes, because you're the ones that are doing the running.

Lenny Murphy: We know. Fine. 10 gifts for each person. Here's my budget. Bring them in, click the button. Look, it's all equal. It's all great.

Karen Lynch: I don't need to worry about it. I know. It's true. Yes, it's true. That's fine. Anyway, we really do have to run. It's getting late. We have two stories we'll share for further reading. There's studies you can read up on them. You know, one of them is about chat GPT, one of them is about synthetic data. Enjoy. We'll leave it at that.

Lenny Murphy: Yeah. Uh, and both are very useful for this whole idea that we're just talking about. So, all right, yeah, uh, Karen, this is a lot of fun.

Karen Lynch: Um, we'll see you next week.

Lenny Murphy: Yes, thanks, everybody.

Karen Lynch: Have a good one, guys.

Lenny Murphy: Bye.

Links from the episode:

Salesforce Signs Definitive Agreement to Acquire Loops 

Salesforce to buy Listen Labs 

Accenture Acquires Sagacity Solutions 

Popular AI leaderboard Arena nearly doubles valuation to $3.1B valuation in 10 months 

micro1 announces commitment to spend $1B on enterprise data acquisition over the next 12 months 

Exciting Opportunity: Moderated interviews with Qualitative Panels 

GWI Launches New Agentic Portfolio 

Numerator: Understand the AI-engaged consumer 

Apple Has Far-Reaching Plans To Block Hundreds Of Programmatic Data Companies From iOS 

IA Bans Fraudsters, Sees Off 'Damaging' Data Proposals 

Too Many Bots in the Kitchen? Williams Sonoma Bans AI Shoppers 

Product.ai Survey Finds Four in Five AI Shoppers Haven't Let an AI Assistant Buy for Them, Nine in 10 Check Its Advice Before Purchase 

AI-Assisted Submissions in Online Research Are Rare and Highly Concentrated but Routinely Approved 

Behavioral History Outperforms Descriptions of the Person for LLM Synthetic Personas 

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

393 articles

author bio

Leonard Murphy

Leonard Murphy

Chief Advisor for Insights and Development at Greenbook

795 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

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...

Special Edition: Career Transitions in Insights with Karen Goldstein and Matt O’Mara
Available on YouTube, Spotify, Apple Podcasts, Amazon Music

The Greenbook Podcast

Special Edition: Career Transitions in Insights with Karen Goldstein and Matt O’Mara

Play episode

Karen Goldstein and Matt O’Mara discuss career transitions, executive hiring, experienced talent, and networking in the insights industry.

Sign Up for
Updates

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

67k+ subscribers