Talk Commerce · 2026-10-30 · 8 min
Key moments - from our scoring
Substance score
43 / 100
Five dimensions, 20 points each
Blue Labs brings a data-driven methodology born from political campaigns to commercial applications across CPG, retail, entertainment, and apparel. Will Sanchez explains that the key distinction lies in targeting persuadable buyers - those who can be swayed to switch brands - rather than propensity buyers who were already likely to purchase. The company's four-step process starts by identifying a brand's core audience through surveys of several thousand consumers, extrapolated across a 200-million-person US consumer file with over 200 variables. Step two uses pre/post design surveys to isolate which messaging actually moves persuadable buyers. Step three develops persona cards detailing what content resonates, where these buyers shop, and their characteristics. Step four deploys matched holdout experiments to measure real lift and determine where to invest or divest marketing spend. Sanchez notes that AI is enabling Blue Labs to move from custom engagements toward packaged offerings, expanding accessibility to mid-market brands. He critiques the retail media landscape for lacking rigorous measurement standards, warning that brands often mistake engagement metrics for actual ROI and miss the opportunity to truly understand their incremental audience before spending scales.
Hosted by Brent W. Peterson. The show scores 61.5/100 on substance across its analysed episodes. See the full show profile.
A propensity buyer is someone already likely to purchase your product and was going to find it anyway; a persuadable buyer is a swing voter in a category who can be influenced to switch brands or try your product for the first time with the right messaging.
Blue Labs conducts pre/post design surveys where they show test messaging to a sample of persuadable audiences and measure whether it shifts their purchase intent, isolating which specific content actually drives conversion rather than just reach.
AI slop refers to basic, non-intricate marketing results where agencies use AI to find lookalike audiences of existing customers and assume they'll convert, missing the opportunity to identify and target truly persuadable incremental buyers.
Retail media lacks independent measurement standards; brands and retailers grade their own tests so they rate themselves highly, masking the lack of real ROI and preventing brands from understanding which consumers are actually incremental.
Blue Labs divides audiences into test and control groups matched on persuasion score and demographics, activates messaging to one group, and measures the actual lift in behavior to determine true ROI and where to invest or divest.
Our reviewer’s read on each dimension, with quotes from the episode.
The persuadable-vs-propensity buyer distinction and the matched-holdout measurement approach are genuinely useful concepts, but the episode is only 8 minutes and ideas are left surface-level with significant filler and repetition. A smart operator walks away with one or two frameworks but no depth.
a swing voter and an incremental buyer, very similar as well, right. So um, the commercial application kind of showed itself pretty obviously
Everyone's grading their own tests, right? So everyone's going to grade themselves with an A or better
Borrowing the swing-voter framework from political campaigning and applying it to retail buyer segmentation is a genuinely fresh lens that most B2B marketers haven't encountered. The retail media 'grading your own tests' critique is sharp but neither idea is pushed to a truly contrarian or actionable conclusion.
we were actually born out of politics, um, where you know, know you have finite budgets. You have one chance to prove it
not around who is buying my product, but who else can I move to the group, right? So propensity versus persuadability
Will Sanchez is a legitimate practitioner with a real methodology and an interesting origin story, but he is a mid-level managing director at a relatively unknown firm, and the conference-booth format pushes the conversation toward vendor pitch rather than practitioner depth.
my title is managing Director for CPG and I work for Blue Labs
we were actually born out of politics, um, where you know, know you have finite budgets. You have one chance to prove it, right
A handful of concrete numbers appear (200 million adult file, 200 variables, 20 million sneaker buyer example, 'couple thousand' surveys), but there are zero named clients, no real campaign ROI figures, no timelines, and the examples remain hypothetical or anecdotal.
we will survey those people, couple thousand and then extrapolate that to our consumer file of about 200 million US adults with over 200 variables
So let's say out of the 200 million there's 20 million, right. That love to run their sneaker buyers
The host provides a serviceable Adidas/Nike framing that helps clarify the concept, but questions are generic conference-booth fare ('How has technology changed all this?' and 'Do you have any hot takes?'), there is zero pushback, and the guest's own 'excellent question, excellent question' sycophancy goes unchallenged.
How has technology changed all this over the past five, ten, however many years? I mean things are changing fast.
Do you have any hot takes for us?
Computed from the transcript - who did the talking, and the words that came up most.
Transcribed and scored by The B2B Podcast Index.
Speaker A: Welcome to this special episode of Talk Shop at Shop Talk. Enjoy our next guest,
Speaker B: Day two of Shop Talk. Last interview of the day
Speaker C: here with
Speaker B: Will. Um, could you start by telling us your name, title and company?
Speaker C: Yeah, definitely. So Will Sanchez. Uh, my title is managing Director for CPG and I work for Blue Labs. Um, yeah, I think that's.
Speaker B: Well, tell us about the uh. Yeah. What does Blue Labs do?
Speaker C: Yeah, I'll tell you a little bit about Blue Labs. We're an interesting house. Uh, we were actually born out of politics, um, where you know, know you have finite budgets. You have one chance to prove it, right. And there's no really like take backs, right. M. So our team really learned very quickly that there's actually within these, you know, within the population of the US that's voting, there's a persuadable group there, right. You could persuade people or swing voter, right. And then very quickly we learned that a swing voter and an incremental buyer, very similar as well, right. So um, the commercial application kind of showed itself pretty obviously. So then now we work with um, various industries. We work from entertainment to government to cpg, uh, apparel retailers. And it's really about not around who is buying my product, but who else can I move to the group, right? So propensity versus persuadability.
Speaker B: Okay. So maybe we have like the die hard Adidas fans, the die hard Nike fans, and then some group of swing buyers in between who can be persuaded either way or flipped from one side to the other.
Speaker C: A hundred percent, 100%. And our methodology across all the industries really stays the same, the same. It's a four step process. Um, you know, the first step, we go ahead and we identify those Adidas buyers, right? So we're going to say, okay, who is in this universe of sneaker buyers, right? Or runners let's say. So we go ahead and we will survey those people, couple thousand and then extrapolate that to our consumer file of about 200 million US adults with over 200 variables. And then from there we can actually start seeing step two, which is step two I think is, is really where it gets really interesting. That's where we can give you the persuadables. So we'll take that group. So let's say out of the 200 million there's 20 million, right. That love to run their sneaker buyers. Great. So let's take that 20 million and let's do a pre and post design and survey, right? So let's see, let's show them the content and then let's not. And then after, let's survey them and see did it move them. Right. Were they willing to buy your product? Was the content enough to say ye, that's something I would buy? Right. And once we can identify those persuadables, we'll give you that group. In step three, which is your Personas or your cohort cards, we could say this is what the content that really moves them, this is the what they're looking for, this is where they shop, etc. Etc. And then step four, which I think is also critical is where we, we put our money where our mouth is and we say let's activate a measure. So we'll do a matched holdout based on persuasion, score, demographics, um, et cetera and then we see where the real lift is. And even if there's no lift, that's a great read as well because now you know where to divest and invest.
Speaker B: Yeah. So let's say that we've identified what, what moves the swing buyer. Where does that message typically get deployed? Is it digital? Is it advertising? Where? Yeah,
Speaker C: excellent question, excellent question. So we work across all mediums. So we work in store. There's some folks that like to test their in store messaging to see what's moving folks or is there specific, um, you know, buy that would move for specific promotion, etc. Um, but you know, at the same time what we do there is we will give you those groups and then activate a measure against them. Right. So to find out within those persuadables what is moving them or what's going to make them buy, we will go into your will, work with you on your content and whatever, you know, whatever you're speaking to them that makes them move and then we will go then and apply that into the um, the deliverable, if that makes sense.
Speaker B: Yeah. How has technology changed all this over the past five, ten, however many years? I mean things are changing fast.
Speaker C: Yeah, things are changing really fast. Um, and I think technology is going to be, you know, especially with the rise of AI, everything is starting to get a little easier for us. There was a lot of manual work. Right. There's a team of data scientists that do all this work for us. Um, but I would say that now with the rise of AI and everything, I could start seeing us packaging these things into more a packaged offer instead of a custom play.
Speaker B: Does that expand the market for you a little bit? Like uh, maybe a little more mid market? I don't know. What are your typical customers like?
Speaker C: I think it definitely expands the market I also think it also gives us other things to measure. Right. So I'll give you an example. I had a customer the other day that was like, I want to apply AI across our huge company that has all these different people that go in there for shipping and receiving. Um, how do I apply AI and see if there's any ROI to that application? Will people use it? Will people use it and like, it? Will people buy more? Those are things that we can go in and test and start holding out groups, seeing which groups are the ones that are doing fine, you know, shipping and, you know, they're using services like FedEx and DHL and all that stuff to figure out, okay, this is your consumer. This is the persuadable. This is what they move for. But now, let's see, is it really the AI that's working for them or are they like, I was fine with the, with, with the original format. Right. So it's, it's, it's, it's interesting. It's opening up different places for us to use our measurement and experimentation.
Speaker B: Yeah. Well, when you talk about consumer applications of AI, I think there's both a lot of opportunity, but also a lot of pushback.
Speaker C: Yeah.
Speaker B: One of the keywords that's, or buzzwords, that's been kind of going around, shop talk, um, I think there's even a panel on it going on right now is AI slope. Yeah. How would you define that term? What is AI slop?
Speaker C: Yeah, there is a lot of AI slop. I think it also falls, that falls into our realm because, you know, when you're working with marketing companies and they all do a great job, but a lot of them are relying on AI. Right. And they're getting, uh, a basic result, but not a really intricate answer on who your customer is or what moves them. Right. They can kind of tell you, you could tell them who the shopper is, they'll look for a lookalike, and then they'll try to figure out if that person will buy. But that person was going to buy anyway. Right. That's a propensity buyer. That's why they're on social media. That's why they're, they're, they're not researching your product, they're just running into it and saying, I'll buy it. Right. Um, and I think a lot of that falls into that AI slot. They're just saying, okay, this will work. That will work for the consumer that you already have. That's not going to work for the consumer that you might be able to get
Speaker B: yeah. Do you have any hot takes for us?
Speaker C: Yeah, yeah. Hot take. Retail. Um, media. Have a hot take on retail media. I think that, um, retail media is great. I think it's all excellent for branding and etc. But, um, there's a huge gap in measurement. Everyone's grading their own tests, right? So everyone's going to grade themselves with an A or better, Right? So I think that, um, I think with retail media, you know, the big hot take there is that I think there's a huge opening for brands and also retailers to step back, right? And before you hit the wall and people wake up to the fact that they're spending money without roi and start measuring what you're doing, start really figuring out who the consumer is instead of just splurting it out there across the universe. Um, and I think you'll get much better results than what you're getting right now. And I'm hearing from a bunch of customers, yeah, I'm getting Lyft. Where's the roi? Right. Where is the actual return on ad spend? Um, and they're missing that. And I think it's because they're missing measurement.
Speaker B: Cool.
Speaker D: You're listening to Talk Commerce. Subscribe and download at Talk.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.
How Performance Creators Beat Meta's New Algorithm with Mal Chia, Managing Director & Co-Founder at Ecom NationThe Partnership Economy · on Holdout testing87 / 100
How to Use AI for Content Without Creating Slop with Eoin Clancy (VP Growth at Airops)The Dave Gerhardt Show · on AI slop87 / 100
Retail Media In Your Ears: Inside Dollar General's 21,000-Store Audio NetworkFuture Commerce · on Retail media measurement83 / 100
Is AI Marketing Helping or Hurting Growth? With Sheila Rondeau Ep. 146Talk to Th3 Doc with Steve Meek · on AI in marketing83 / 100
How to Build a Brand People Love | Sarah Holt on Marketing, Leadership & AIThe Places We'll Go Marketing Show · on AI in marketing82 / 100
CTV Measurement Unpacked: Incrementality, Attribution & Holdout Testing | Tom RathboneAlways Be Testing · on Holdout testing81 / 100
