What are the current and upcoming trends in the CPG data space? In this episode, Ian sits down with Sean Cline, Industry Principal for Retail and CPG at Seek, to discuss in depth with Peter V.S. Bond, Partnership Acceleration Lead at Flywheel. Peter is the founder and co-host of The CPG Guys podcast, where he regularly covers the latest trends, insights, and best practices in the CPG and retail space.
What are the current and upcoming trends in the CPG data space? In this episode, Ian sits down with Sean Cline, Industry Principal for Retail and CPG at Seek, to discuss in depth with Peter V.S. Bond, Partnership Acceleration Lead at Flywheel. Peter is the founder and co-host of The CPG Guys podcast, where he regularly covers the latest trends, insights, and best practices in the CPG and retail space. This conversation covers the impact of generative AI, augmented and virtual reality, and the increased adoption of clean rooms.
---------
Key Quotes:
“A lot of people think that you need to own the data. We live in a world with clean rooms, with disparate data sets, and there is no need to own data en masse for every data component that you need.” - Peter
“That's the power of big data. I can create audiences that are meaningful to me, and I can treat each of those audiences fundamentally differently, and I can measure the performance of each.” - Peter
---------
Show Timestamps:
(02:33) What is Grocery Shop?
(05:55) AI takeaways from Grocery Shop
(07:50) Where is AI having the most impact?
(10:48) Augmented and virtual reality in CPG
(15:57) Changes in data through Peter’s career
(21:29) Can everyone leverage data?
(25:57) The gap between business and data experts
(30:34) The increasing adoption of clean rooms
(33:38) Upcoming trends that will have impact
(38:35) Why media retail networks are exciting
(44:55) What data is not collected that could improve CPG performance?
--------
Sponsor:
This podcast is proudly sponsored by Seek, the leader in cloud-based creation and delivery of industry-focused insights.
--------
Links:
Connect with Peter V.S. Bond on LinkedIn
Connect with Sean Cline on LinkedIn
Connect with Host, Ian Cook, on LinkedIn
Follow Seek on LinkedIn
Ian Cook: [00:00:00] Some days it feels like we're drowning in data, but data isn't knowledge. Data without context or the expertise to understand it is just something taking up space in your warehouse. That's where being insights driven comes in. Insights driven organizations focus on generating actionable insights rather than just collecting and munging data.
Insights drive better decisions. Welcome to the Insights Factor.
Thank you to all the listeners joining us today. It's going to be a fantastic episode. Not only am I joined by a guest, I have my frequent, um, co host for those who've listened frequently. I've got Sean Klein, our industry principal here at Seek. Uh, you can definitely tell where he works if you're on the video feed, but the big guest here today.
Doesn't really need an introduction, but I'm going to do it anyway, just to make him feel self conscious about it. Technically, by day, he is the Partnership Acceleration Lead [00:01:00] at Flywheel, but my guess is you're going to have heard of him and have listened to him on the illustrious podcast, The CPG Guys.
And I have to start out simply by saying a thank you for CPG Guys. As I've mentioned almost every episode, I come to this in a very different way. I'm from a technical background, a developer background, doing machine learning, AI, data science kind of work, and have had the fortune, good fortune, of working with people in the CPG industry to learn this.
So, part of this podcast is me learning, um, and figuring out what's going on, but I, with people like Sean Klein. and being able to get us contacts like the, uh, Mr. Peter V. S. Bond. So, uh, I am thrilled to have you here. Thank you so much for joining me.
Peter V.S. Bond: It's a pleasure, Ian. Great, great to join you and Sean. What a wonderful day to do it, too. Yeah. As the crisp air in New England starts to permeate the world we live in.
Ian Cook: Fantastic. The reason we actually have Peter with us today is not because of me, not because of my reach out, but [00:02:00] because we're lucky enough to have Sean at our company.
And I believe the two of you ran into each other in an industry event recently. Is that right, Sean?
Sean Cline: Uh, Peter, we seem to see each other at grocery shop and shop talk only like Vegas and, and we're together.
Peter V.S. Bond: And, and we're talking about, I thought the Vegas rules were everything that happens in Vegas stays in Vegas.
And here we are getting together after we're making an exception.
Sean Cline: That's as far as it goes. That's as far as it goes. Right. And you're significantly taller in person than I thought you would be. Now, granted, I'm significantly shorter in person too. So, um, but yeah, it seems like we only see each other at these, these events, but happy to, happy to have you here.
Really super excited. I've got a couple of questions. Well, first, let me start by saying it's 80 degrees here in Bentonville. And I think you said you've got fall weather happening, so we're still in shorts and sweaty. So,
Peter V.S. Bond: um, I was 37 when I walked out the door this morning to take my daughter to school and it was, it was, it was nippy.
We'll just say that. That's a wonderful
Ian Cook: mid fifties here in Pittsburgh. But Sean, before you get into the specific [00:03:00] questions, tell me what the most recent event was and give our audiences a quick notion of what that industry event is.
Sean Cline: Absolutely. I'll take a swing at this and Peter, please pipe in with your opinions as well.
Uh, grocery shop was the most recent, uh, it was in September. Uh, that is where, uh, Seek launched our, our new product. Uh, but the event itself is very focused around the retail industry, specifically, uh, Um, technology and technology within the retail industry. So that could range from autonomous vehicles to smart shopping carts, which were kind of weird.
They automatically scan things as you put things in the cart, uh, and whatnot. I don't know, Peter thoughts?
Peter V.S. Bond: Yeah, I so grocery shop was created in 2018 is an offshoot of ShopTalk, which was more focused broadly on retail. So it was more encompassing of big box retailers, department stores, you name it. And there was a calling that they needed something a little more [00:04:00] specific to to grocery industry.
So they ended up creating grocery shops. So I think there were a little over 5, 000 attendees this year, up from year one, which was 2000 back in 2018 pre pandemic. And what I would say it was decidedly grocery, a lot of brands present, not just brands on stage, a lot of brand people. I know quite a number of manufacturers.
Uh, in the sales and e com organizations brought their whole teams there to learn, to interact, to connect with one another and, and to see what's going on in the industry. So I would say when you compare it to some more legacy industry events, I'm talking about the Food Marketing Institute, FMI. Midwinter meeting.
I'd say the median age, uh, for, for grocery shop is about 20 years younger than the FMI event. So, it's the guys are helping to bring that down. That's nice. We, well, I'm bringing, I'm, I'm, I'm, I'm definitely not in the standard deviation. [00:05:00] Uh, I'm definitely on the upside, but a, a younger crew and, uh, to Sean's point, very focused on technology enablement.
Digital Path to Purchase, Omnichannel, but I would say that what was interesting about this year was, uh, sure, the, you know, like the, I think of Wheel of Fortune, the final, the final puzzle when they do the RSTL& E where they give you those letters. Like, of course, everybody's talking about Uh, generative AI.
That was up there. But I think with the return to stores and people shopping in retail, there was definitely a focus on getting things right at retail. Supply chain was a big issue. So it wasn't as, um, it wasn't as digitally, uh, monolithic as it had been the last couple of years during the pandemic. But it's, it's definitely still predominant, but some broader issues that all retailers are starting to deal with.
But a great conference. It's One of my two favorites, the other being, being ShopTalk, and I wouldn't mess it. It was [00:06:00] great to see Sean there.
Ian Cook: Oh, that's terrific. Well, it's good that we can be at those. SEEK is, is, uh, relatively new, but the people we have are not, so it's kind of fun. Wish I'd gotten a chance to go out there maybe next time, uh, to see how this whole thing works, because I think it's fascinating getting to talk to these people.
But, um, Sean, it looked like, uh, you had another question.
Sean Cline: Yeah, I was just thinking, I feel like you've answered this already, but what would you say the biggest takeaway for you was from, from Grocery Shop?
Peter V.S. Bond: Yeah, for me, it's, it's that retailers are trying to keep pace and embrace a lot of these new technologies and capabilities.
I think there were no fewer than three sessions devoted to the topic of AI. But when talking to the leaders and asking the question, Uh, are you building or you're buying? The answer was decidedly, well, I'm buying because if I try to build, there aren't enough people out there in the industry with the subject matter expertise [00:07:00] around developing AI applications to satisfy what would be the demand of every single major manufacturer in the industry.
So, they're relying on third parties like Microsoft and its investment in OpenAI or Google and BARD. to start getting their feet wet. They are still focused on what I consider fundamental activities. So if you think about you're an omni channel retailer and you're selling across multiple different platforms, just the thought of like trying to create content that is compliant with a specific retailer.
And then you have, you know, several hundred retailers you're trying to do that with. That's just not. Humanly scalable without an armada of people. So thinking about using AI to just modify. textual content or even creating image iterations that are size compliant with the product detail page of this retailer versus that retailer.
That's kind of like the entry point where I think most [00:08:00] people are going. Um, not much more sophisticated than that, but a lot of promise and a lot of aspirations. That, that was my biggest takeaway from this event. And a
Ian Cook: question on the generative AI. Where specifically do you see most of the impact? Is this going to be between the brand and the retailer, say, operational, or is it more retailer to customer, and it'll be the interface and the way the customer interacts with something?
What were people talking about in the ways that generative AI for this world starts showing up?
Peter V.S. Bond: So let's talk the B2B first. I think the ability to improve the content on a product display page to make your product detail page rather than make your product more discoverable is a noble and meaningful pursuit.
So I think it's going to do that on the B2C side. That's the really interesting one. I think it's going to have a fundamental effect upon how people conduct search. If you think how search is developed, it's very much a [00:09:00] vertical experience, and this is what I mean. Try to picture, uh, the, uh, aerial view of a grocery store.
You walk in and you're on the perimeter, and when you think of something, you dart down into the aisle, you go get it, you come back out to the perimeter, you go two more aisles over, you get something else, and that's a very vertical search because you're searching item by item by item. And that's how Amazon works.
Amazon is a, is a, is a, uh, has an endless aisle, but it's all about eaches. Most people come to Amazon, they know what they want, they type in the search term, they grab what they want, they put it in the basket, they check out, and they're done. But I'd ask you this. Do either one of you think about shopping on Amazon for the 4th of July party backyard barbecue?
No, because... Ultimately, it's a lot easier for you to walk into your local grocery store or Target store and go to the section of the store that they've devoted to the Fourth of July, where you can throw everything into the basket. It's much more time saving. So that's where generative AI is having a [00:10:00] profound effect.
Look at what Instacart has done with Ask Instacart. Instead of me saying, I want chicken, I want barbecue sauce, I want charcoal, you say, I'm throwing a Fourth of July party. And then what? Instacart does is it comes back with a list of curated items that are relevant to your need state and the problem you're trying to solve, so you don't have to sit there and enter in all those items.
It should get, and if it's, if it's learned who you are, it understands when it comes back and it shows you buns and, and you're living a gluten free lifestyle, the buns it's going to show you are gluten free. That's where generative AI is going to have a tremendous impact on our B2C aspect, and brands are going to have to figure out how they win in that kind of a situation and how they insert themselves into the process.
Ian Cook: That seems to be where you, like you said, the becoming more discoverable, making the documentation of the description hit. on [00:11:00] those things that people are going to be feeding into their generative AI system in order to understand better. I wonder, because I wasn't at Grocery Shop, obviously, is there anything here where that starts playing into sort of the artificial reality or augmented reality and the virtual reality?
Do you think people will be walking around with their Apple visions and they see the list of recipes that they want to make for the week or the 4th of July party and they can check off and just walk through the space and say like, all right, I can either pick it up either. At the store, just so I know where to go, because sometimes I'm at a store that I don't know the layout that well, or completely virtually, and I just want to have that experience of walking through the process, which is something that I think we've struggled with when we think about like the Amazon, or Instacart, or any grocery shopping online, is you don't have that sort of um, discovery experience where you're walking down the aisle going, Oh, that's right.
I would love some pinto beans with whatever and you, you know, run because you run across it.
Peter V.S. Bond: So I think there are a couple things here that we have to unpack. What is the use case for [00:12:00] augmented or virtual reality? To me, it's about bringing to life experiences that drive conversion. So when I think about apparel, I want to know does this shirt that I'm buying look good on me.
So to be able to create that kind of environment, L'Oreal has done a really great job in terms of all their products and trying to create that Warby Parker with putting glasses on. Those are the kind of things where augmented reality can be really powerful. Household. You want to buy a sofa. You want to know what it looks like in your world.
Those are the kind of things where I think it has tremendous value. I'm not so sure about necessarily I want to shop a virtual store because it doesn't really, I don't really need that experience. Maybe with a recipe so I can see how a product is prepared and I can actually watch someone shopping it and it's more than just.
The straight up video, it's giving me a little more context. You know, like when my wife says, my wife is a baker [00:13:00] by, by skill. And I don't say that she's really skilled at it. I'm saying that she doesn't know how to cook. She knows how to bake. She knows how to bake because it is precise instructions. This amount of ingredient in this kind of a pan for this number of minutes.
If you, if the recipe says, Brown the chicken on both sides. That is way too much for her. She's just like, no, no, no. How many minutes? What temperature setting do I need to do that? That's where augmented reality can actually help someone who's, shall we say, culinary challenged in terms of their, their cooking skills.
I love my wife. I'm just, you know, I'm just
Ian Cook: teasing her. No, I've always heard cooking isn't art. Baking is science, right? Yeah, exactly, exactly. Browning is a matter of taste. The exact amount of baking soda that you want in a recipe is, is something that you can measure out easily. So it's, that's really fascinating.
It's an interesting way to think about it. [00:14:00] So we don't see that as much. Is there any possible use case for the augmented or virtual reality for, say, the people who are doing the work at the grocery store or even the people you shop for? with the brands like checking to see as they walk through to try to do stock and understand like how the their items are being purchased.
Peter V.S. Bond: Sure. I mean, if it's, I think about the original Google Lens device when it would pop information up on, uh, and it would only be visible by the wearer. So if I'm shopping the store and I'm, if I'm using a store as the fulfillment center and then I need to pick items, it could. very well, create that environment where I'm able to tick my list off very quickly, and it can facilitate a more expedient process for fulfilling an order.
That, I think, has tremendous value. Or, uh, it can highlight a great, great tool that Instacart does. They have the little carrot function where, as you're walking through the store, the app will tell you where the product [00:15:00] is on the shelf, because when you're trying to find that tiny little bottle of hot sauce, Among 600 hot sauce bottles, it'll say it was right down there on the fourth aisle.
So yeah, it can, it can add some value to the, to the in store shopping process for both, both someone who is trying to fill the order, or even for an individual shopper who's trying to find the product. Cause when I, I know where the core items in my basket are. for the things I'm buying repeatedly, but if I need to find, um, you know, cream of tartar because I'm creating some kind of a souffle, then, uh, it's gonna be, I know it's in the spice section, but I have no idea where it is.
So, that can be useful. Exactly.
Ian Cook: Well, even before the world gets to all this augmented reality, virtual reality, generative AI being core to everything that's going on, They're going to need data, and that's one of the big things that we think about and talk about here at the Insights Factory is this kind of data.
I'd like to go back a little bit in your history.
Peter V.S. Bond: Are you someone who's used data a lot? Are we going back by, uh, you know, are we going back in time [00:16:00] by Hot Tub or DeLorean? Because I kind of need to know.
Sean Cline: It's a room for me, the
Peter V.S. Bond: DeLorean. Yeah, it's flux capacitor. That's all I have to say,
Ian Cook: Chad. That is much more my speed.
I know I'm aware of the hot tubs I travel, but, uh, I've, you know, never been that comfortable with hot tubs, so. Uh, thinking back, is this Is data something that was prominent in the way you did work before? Has it grown over time?
Peter V.S. Bond: Oh my god, yeah. So when I think about entering the CPG space in the early 90s, after I got out of business school, big data was centered primarily around uh, transactional volume from retailers.
So the two big players in this space, IRI, now known as Cercana, and Nielsen, now known as Nielsen IQ, they, they, they made their claim to fame by procuring from retailers store level transaction data, [00:17:00] aggregating it, Um, cleaning it up and organizing it and then reselling it to manufacturers so they could do basic functions like understand what's my share in the category.
What, uh, did, did I grow the business? How much of my volume could be attributed to, uh, incremental sales driven by a display or a feature in the weekly circular? And that was, that was crunching a lot of data. And when I entered, for the most part, These two players would sample about 10 percent of the stores in a retailer's chain and then project out the sales for the entire chain.
Why? Because they didn't have the functional capacity to process that much data. So
Ian Cook: the limitation was on processing, not collection. They could sample it from anywhere, but they couldn't do anything with that mountain of data in any sort of reasonable amount of time. Exactly.
Peter V.S. Bond: And fast forward now and we're talking about massive amounts [00:18:00] of data that not only take into account the products in the store and the causal conditions, whether it was on display or things, but actually who the consumers are, who longitudinally connecting transactions.
At a household level so that you can understand the purchasing behavior over time of individual consumers, that unlocks a tremendous amount of capability. Prior to that, IRI and Nielsen were maintaining what they called consumer panels. They were a group of about 100, 000 households that they would Select based upon their conformity to the demographic distribution that the Bureau of Labor Statistics and the Census told them were, were ultimately representative of the United States.
And the problem when you do that is you're, again, projecting, but when you want to try and look at a category like shoe polish and the purchasing behavior of 100, 000 households in shoe polish, [00:19:00] When you go below the category level, even the brand level, the statistical sample size is so insignificant that it's not measurable or actionable information.
Fast forward now, Kroger tracks 60 million households. It has Two and a half years of longitudinal data loaded into its data asset. And it's able to understand at an item level so many components of the transaction and build what I would call our propensity signals. And propensity signals are, like, demographics are useless in big data.
They're all that we had for a long time. But the way I best describe that is If you're, if you're Diet Coke and you're trying to sell more of your brand, are you more interested in targeting women who are 18 to 24? Or is it more efficient to target people you know are heavy purchasers of diet cola or diet soft [00:20:00] drink products?
Because there are a lot of 18 to 24 year old women who don't drink soda and never will. And so it's a wasted effort to try and advertise to them. When you can process this amount of data and I can sit there and use those signals to create. Enormous audiences that even as I start to apply all these filters to them, they're still sizable enough that I can create an audience, I can send an impression to them in the form of different things, digital advertisements, what have you, a direct mail piece, uh, all of this, and I can actually measure the performance of that and, and attribute those sales and understand how much of that, and then even hold a control group.
So I can say, what was the incrementality of the test group that got this versus the control group that didn't? That's fundamentally changing how brands are able to advertise and justify to, to their [00:21:00] investors, their seniors, and the company, the board of directors, that in the, in the demise of, of more traditional mass advertising mechanisms like linear television and print, print media, which are declining at a, Precipitous rate.
These are mechanisms that are not only replacing it, but are much better because I can actually measure performance. That's the power of big data. Sampling is both
Ian Cook: statistically powerful, and materially powerful. So even if you could get a statistical, uh, uh, an appropriately significant statistical sample, it might not be a big enough group.
Now we have both, and you can take those actions. Yep, but what I wonder is, on the human side, has it, have the skills to understand and work with this data kept up the demand for it? Or are we in a world where there's just all of this data and we're still trying to get it to the point where it's something that everybody in the system can be using and take action
Peter V.S. Bond: on?
So it's very important [00:22:00] For advertisers, they focus on the outcomes they're looking to achieve and that they use the data tools and the metrics that are reflective of what your outcome is. Just because you can measure it doesn't mean you should. measure it. Like if your goal is to launch a new product, then return on advertising spend for a new product launch is not your primary key performance indicator.
It's really about impressions. I'm trying to create awareness of a product. If I have a more mature brand, right, Then these concepts of return on investment or return on advertising spend are much more, uh, much, much more meaningful because you don't have to create this awareness. You're not spending this stupid money on making people aware that your product even exists.
You're just telling people who either are loyal buyers of your product or occasional buyers of your product or buying categories that are adjacent, [00:23:00] here are my products and I've got a a fundamental offer that's designed specifically for you. So if, if Sean here is, uh, is a, an avid loyal buyer of my brand, my goal is to get him to expand his consumption.
So the offer that I might give him, let me say, may say he spends 30 a month on my brand. I may send him a specific offer that says, you know, Sean, if you spend 40 over the next month on my brand, you're going to get An extra 2 back. Whereas with you, Ian, you've, you've never bought it before. My goal is to just get you to buy it once.
So I might give you a very rich offer, but the threshold that I'm trying to get you to achieve is not 40 like it is for Sean. It's just by one unit. That's the power of big data. I can create audiences that are meaningful to me and I can treat each of those audiences fundamentally differently and I can measure the performance of each and so that is where big data [00:24:00] is great.
Now to your point on Is it, is the capability there? Yeah, I mean, the great thing about it is, I would, I would argue that probably the foremost authority in customer data science is Don Humby's Tesco, uh, pardon me, Tesco's Don Humby division that, uh, that became famous for driving data for The club card at Tesco, and then they helped usher in customer data science at Kroger.
There are enough people in the ecosystem now that are expert data scientists that permeate consumer package goods companies, data and insights providers. It's really just about understanding how to build audiences. How to deliver a particular experience, and then exactly how to measure it on the back end.
Things like understanding, if I'm trying to measure the incrementality of a box of cereal, right, it's fundamentally different than a Samsung television. Why? Because I'm buying serial once a month. [00:25:00] I'm buying a TV once every three years. So when I look at the, the post purchase period in which I'm going to pull forward and understand any consumption patterns, a tele, you know, how I'm going to do that for a television is fundamentally different than how I'm going to do it for serial because I'm just, I'm out of the buying cycle.
So you just have to have experts that understand how to use data and understand the nuances of not only how to measure, But purchase life cycle of particular products to help get you there. This excites me
Ian Cook: because this is the exact gap that I think we're trying to address here. And that really interests me is obviously there are data scientists and the growth of data science and machine learning has blossomed.
I've been a beneficiary of that. Yeah. When we get into these businesses, a lot of the people we talk to that want the data, that want to do something with it are not necessarily the ones who are experts at the data science or figuring out even how to get at the data. So there's this, to my experience, and this is what I want to gut check [00:26:00] with you is people who can back the data up, like it's a huge truck and say, here, it's all in front of you.
And then the people who are like, I've got to do something to figure out how to get. Me or Sean, like, whether is Sean the buyer that I need to expand or is Ian the person I need to get to try it. So, living in that kind of gap, it still seems to me that there might be some, um, missing connection there that, you know, the whole business wants to be quote unquote data driven, and it, but you put tons of data in front of someone, they might not be the, they're the business experts, like you said, they're the experts who are like, I know what to do with this, But I'm not sure how to get this out
Peter V.S. Bond: of the data.
So that's where outcomes become important. So if I'm a, if I manufacture sunscreen, then the kind of data points I need is I need to understand in real time, what is the UV index across major metropolitan areas? And do I have a mechanism so that when it spikes above a certain level, it does a push notification to all the users of my [00:27:00] retail mobile app that the sun, the, the UV index is above a certain range.
We highly recommend you get sunscreen on today. And oh, by the way, for the one, for the, for those people receiving the message that are highly respondent to offers, those people and only those people, Get an offer to buy it at a discount and that's enough to drive them. That's really where data and understanding how to apply that become very important.
But that also doesn't mean necessarily that that retailer needs to go out and quote unquote, build or buy their own UV index measurement tool. A lot of, a lot of people think that you need to own the data. We live in a world with clean rooms, with disparate data sets. There is no need to own data en masse for every data, every data component that you need.
That is why CleanRooms exists. CleanRooms are the ability to, [00:28:00] to, in a, in a privacy safe environment, match data sets together and then enrich the data on both sides of the equation to get at that without me having to necessarily know, uh, the user information that's matching the data set. Uh, from, from the third party provider to myself.
I just want the outcome. I don't need to own the data. Once you get over that hurdle and you understand that there are so many other data sets that you can access that you don't necessarily have to own, you just have to be able to match against, now suddenly you unlock the capability. And again, focusing on the outcomes is what's going to let you succeed in using that data to drive desirable outcomes.
Sean Cline: Does it go as far as minority report?
Peter V.S. Bond: I gotta know. Uh, it's not quite as bad as that, but, uh, but yeah, it's, it's definitely in there. Now, now my mind's going to be running on Tom Cruise all day. Thank you.
Ian Cook: Right. Yeah, you're welcome. Sean, what is the pre crime of the CPG world? What is something you'd have to run and jump into [00:29:00] somebody's house before they commit the, the terrible sin?
Yeah, I was, I was thinking
Sean Cline: more along the lines of, uh, you know, Sean Klein, 50 year old guy walking into a Walmart. Store 5260 in Rogers, Arkansas. I love shoes. It's going to serve me up things as I'm walking in those doors because it knows I love shoes.
Peter V.S. Bond: It's not any different than an Instagram feed. It figures out what I like across different platforms and it's sending me out things.
It is. I know they are. I know they're, but dammit, I, I'm always buying things from them. I, they know, they know me, but it's interesting. I'm buying things that I see on Instagram. I'm not necessarily buying them on Instagram. Uh, and, and I, and what's happening is I tend to find something and then I want to go to a platform that I trust where I've already entered in all of my credit card and mailing information.
And more often than not, I can find. What I saw on Instagram on that platform, it's probably cheaper and I'm going to get it a lot faster. So, [00:30:00] we now live in a world where we advertise on one platform, and at the end of the day, can you measure In a omni channel cross platform method to understand, like, there are a lot of people that talk about they advertise on, on the advertising platform of that big retailer out of Seattle, Washington, and when they stop advertising, they see their brick and mortar sales.
In markets decline, and that's because it has an effect on how we shop in general, not just how we shop on that retail platform. That's why data is so important. When
Ian Cook: you are talking about these things, like the data room, like enriching data, like not needing to own data, do you find yourself at the leading edge of this and being a thought leader in this?
Or are a lot of companies taking this step towards what you're talking about?
Peter V.S. Bond: So clean rooms, I think, in the last year and a half have been more broadly embraced by most major companies. They use third [00:31:00] parties, uh, to do that. Some retailers have their own clean rooms. Certainly, Amazon's most notable for Amazon Marketing Cloud, which is a huge clean room.
They just introduced at their unboxed event two weeks ago, the Amazon Publisher Cloud, which is a clean room for publishers. Um, Kroger has one. Walmart has one, they all have one. Um, but then there are third parties, companies like Snowflake or, uh, a couple of others that are that are essentially Habu sits on top of Snowflake's data set.
Now, there are lots of them that use these and really what it's about is just mechanisms to share data in a privacy safe environment. So, I'm more interested in not who it is, but you need to use a clean room to make. Other data sets more accessible to yours and enrich your data set for the purpose of driving outcomes.
This is very
Ian Cook: much [00:32:00] where we're aiming our tool Seek Rich. It's not a clean room, but it is a tool to provide that kind of enriched result. The other thing is that you don't have to own the data. We can be holding the data for you. Do you find that there's a difference in the size of CPGs? Are the smaller ones taking this on because it's more cost effective?
Or is this the purview of the larger CPGs because they've been there and realize their efficiencies to be had to making these sort of technological improvements?
Peter V.S. Bond: A year and a half ago, I would have said that the lion's share of CPGs engaging on this, you'd find on the top 100 list, I think it has moved, uh, down the, down the chain to mid size manufacturers, it's not just the big ones that we're all necessarily familiar with.
The digitally native ones, they're basically just using the, the marketplace systems that, that already exist. I made mention of Amazon Marketing Cloud, they're availing themselves of those kinds of [00:33:00] solutions to do it, but they are, they are developing the skill set. Um, I would say that they tend to focus, the smaller you get, the more you tend to focus on one, maybe two retail platforms and then using whatever tools they have as data sets.
When you're a larger consumer packaged goods company, multiple brands, multiple lines, then you're Then you're much more focused on not necessarily using a retailer's platform, but using a tool that is more, um, more reflective of the data sets that you, you're interested in. And that's really where I see them going.
Oh,
Ian Cook: excellent. What are some of the gaps that still exist that you think companies could be taking advantage of? What are things that you would say? Uh, everybody could be benefiting from doing more of this on top of the things you've already mentioned. Are there things that are, is there, uh, new trends or new items that you think, like, this is really going to change things?
Generative AI being, of course, one of
Peter V.S. Bond: them. Yeah, generative, generative [00:34:00] AI is going to have a fundamental impact upon, uh, I mean, e, when I, as mentioned when I started, This conversation with you. I work for Flywheel Digital, which is a managed services company that helps companies with search engine marketing and search engine optimization.
If you think about what drives search engine optimization, it's everything on the product detail page, the description of the product, the number of images. that are selected, the copy that's on the page, even ratings and reviews. You know, one thing that what I think what Amazon did, it got a lot of negative press when it was first released, but I actually think it's quite ingenious.
They talked about ratings and reviews and they were using generative AI to summarize key themes found across reviews. In the old days, the way if you went to a product detail page on on most retailer sites, they would They would say, here's the most popular positive review, and here's the most popular negative review.
Now, when [00:35:00] Fat Review didn't encompass all of the aspects that consumers consistently referred to as being positive or negative, then you were kind of missing the boat there. The ability to kind of summarize that into something in a, in, in, in digestible English or whatever language the right, the reviews are in so that I can get.
Just to read that one thing and that's what that's fundamentally game changing for me. That's, that's, that's natural language processing at its best. So I think AI, AI is a really big thing. Privacy is still going to be a big issue. You know, we talked about GDPR in Europe, consumer, uh, the California Consumer Protection Act, CCPA.
There are probably 11 other states right now passing or have passed similar types of privacy laws. Everybody's always concerned about, yeah, I want access to that data and I'll mingle it with mine, but I need to make sure that, one, I'm not giving them any of my Personally identifiable information and God forbid they give me any of theirs because I don't [00:36:00] want to be, I don't, then I got to call in the lawyers and that's when you start racking up the big bills.
Sean Cline: Oh, that's where those, those data clean rooms come into play too, right? Exactly. PII doesn't, doesn't sneak through. Never does. You just know a 50 year old guy who lives in Rogers, Arkansas, not Sean Klein, 50 year olds.
Peter V.S. Bond: Exactly. Lots of protection. So I think that, that's a big trend. I think, um, supply chain and integration into the omni channel experience is something that still needs a lot of work, and there's huge opportunities, like, if you think of e commerce and the journey as ending when someone clicks the buy button, then you've just missed it.
Thanks. Thanks. Comprehensively missed the point. From my perspective, that's just when the journey begins, because from that moment forward, your goal is to satisfy the customer and not make them regret they bought that product from you, you need to get the right product. They, well, this
Ian Cook: is the Instagram thing you were talking about.
Exactly.
Peter V.S. Bond: You [00:37:00] want to get it to them quickly. You want to get it to them in the condition, the best possible condition. You want the way you represented it on the product page and through the videos to actually, you know, It's the worst reviews that where the worst reviews typically come in is when they didn't actually accurately represent the scale of the product in the images.
And so you think you're getting this big thing that you're picking up and you find out it's like that big it does what it said it did. But if you don't accurately reflect the scale of the product, you're going to disappoint consumers. So that's why having supply chain at the table to understand. every aspect of what is needed to deliver on that customer experience and making sure that anyone who has beginning to end an investment in that process that their performance is graded based upon the success of your initiative that's when everybody that's when everybody [00:38:00] performs at their best because they have they have table stakes right there they're in the game and so for me That's, data is going to be the key enabler of all this.
Uh, think about, like I said, you know, there, there are arguably, what, about 600 retail media platforms and growing, right? So, your ability to create content for 600 different platforms and have it be relevant for one versus the other versus the other, and that's, that's, That's data, that's insights, that's understanding the other side of the table and pushing the right content to the right person at the right time for the right experience.
Ian Cook: So this is where I'm going to take advantage of the expertise I've got here in front of me with both Sean and Peter. I understand retail media networks, very exciting, kind of new. Give me a one on one on that and why they're so exciting.
Peter V.S. Bond: Sure. Uh, so retail media. Retail media is media. It's just media that [00:39:00] is, uh, that is published by retailers.
So if we think of traditional media, linear television, advertising on local TV, the Super Bowl, print media, the Sunday newspaper, that's all just media. What retailers are doing is, is they have the ability to communicate with They're shoppers. How do they do that? They do it a number of different ways. If we go back 20 years, the way that personalized retail media evolved is the delivery of, say, using the longitudinal purchasing history of consumers to curate and select coupons that were relevant for Sean.
So Sean would, let's say the three of us got a mailer from Retailer A, and we each had 14 coupons in it. The most effective way to do this is Using our longitudinal purchasing data to, Ian, you're going to get 14 coupons, Sean's going to get 14, I'm going to get 14, and the commonality between those is only going to be [00:40:00] dictated by the, by the, the commonality of us actually buying particular products.
For the most part, we should have fundamentally different sets of products. 'cause we're different kinds of people. You're gonna probably see some common things like bananas and milk . Fundamentally, it's gonna be different, right? That, that's really how it evolved. Now it's okay with the deprecation of we're what, 18 months away from a cookie list world where?
Where already iOS, Firefox, and Safari have ended. The ability to track users on their platforms. And Google has announced 18 months, they're going to stop letting that happen on Chrome. Right? Well, retailers... They're not, they're not fundamentally affected by that because when people come to their site, they're always logging in to, to be able to see the deals they get.
They've got all of that information. So they're able to then serve up advertising both on their site and off their site, tie that back to the consumer, and then [00:41:00] actually measure the performance of that advertisement. So they could, they combined with They combine with programmatic advertisers, so third party demand side platforms like the Trade Desk or LiveRamp, right, to send out advertising.
They connect with social media platforms like Pinterest and, uh, and, and Instagram and TikTok, but they also have their own assets, their mobile app, their site. And so, up and down the funnel, they can, they can target consumers to, for particular advertising treatments and actually measure the performance of that.
That's the real value. Now, why is it appealing to retailers? If you're a grocery retailer and you've been operating over 100 years on a net 2 percent margin and suddenly you're presented with a world where you're getting 60, 70 percent gross margin on advertising, uh, yeah, I want, give me some of that money.
That is a sea change. It's fundamentally different. So that's why you see every, you know, I like to joke the dry cleaner down my street. [00:42:00] She's got a bulletin board and in her lobby, she calls it a retail media platform. She puts, she puts other business cards up there and she's trying to get, she's trying to get referral fees from that.
But that's basically what, what retail media is. It's just media that is controlled and owned by the retails. But if you think about in store, that's, that's the real. For brick and mortar retailers, the ability to put screens on end caps, to advertise at the checkout kiosk, even to do non digital things like print, uh, shelf tag extenders, so it goes below the The barcode and the price and the description little thing drops down and it's got a nice little promotion saying, Hey, this, this item is on deal, or it's compliant with these different types of health lifestyles.
That is, that's, that's retail media because the retailer owns the space and dictates what can go there.
Ian Cook: A long time ago, wasn't this the way the, the PA system in grocery stores was used as like, you know, shoppers, there was, [00:43:00] there's a special
Peter V.S. Bond: on Kmart
Sean Cline: shoppers, right? That's a blue light
Peter V.S. Bond: special, right? Yeah.
You would. So right now you've got 7 11 announcing that. That, that, um, in store radio is a core component of their retail media platform. Same with Walmart. All of those are mechanisms. Now, I think what you have to understand is when you think about in store media and the experience, it's going to be, at least for right now, fundamentally different than what you see through digital advertising on site or programmatically.
Those are highly personalized. I know who you are. I can deliver the right experience. So it goes from being a personalized experience to being a meaningful experience. You try to advertise your brand, but you, you know, like when you and I are both standing and staring at the same NCAP, it can't personalize the content for you versus me.
It just has to deliver [00:44:00] content that hopefully is meaningful enough that one or both of us will actually want to pick that product up and put it in our basket.
Ian Cook: Right, so it's not just creepy like Minority Report that it knows Ian Cook's standing there. No, no. It knows that I need the gluten free rolls for my 4th of
Peter V.S. Bond: July...
Well, it's... I think many of us are familiar with the story of Target and how a father once found out that his daughter was pregnant because Target sent him a message saying we noticed you recently bought a pregnancy test and we'd love to be able to talk to you about prenatal vitamins and that's the kind of tear point creepiness that, uh, that kind of warning most retailers and brands also don't want to put it.
the consumer in that condition to blame them for, for those kinds of adverse situations.
Ian Cook: All right, Peter, I'm going to give you a magic wand. There's tons of data out there, but what... George Clooney! [00:45:00]
Peter V.S. Bond: Oh, all right, then we're done.
Ian Cook: That's it, folks. There's an amazing amount of data out there. But what is still something you don't see collected, or at least that is not being utilized as much as you think it could be, uh, for improving the way CPGs perform and the kinds of decisions they can make?
Peter V.S. Bond: I think the challenge is retail media is really powerful because it gives you very precise purchase behavior at a consumer level. What's challenging is that even among, if you look at a retailer like Kroger, and I know this because I, I worked with them for many years, even among their top 5 percent of most loyal shoppers, 43 cents out of every dollar they spent on consumer packaged goods was spent outside of Kroger.
So my ability to understand. The omni channel perspective, and in this idea of what I've heard [00:46:00] called the clean room of clean rooms, where I take all the different clean rooms from all my different retailers and I link it all together, like, yeah, great, great pipe dream retailers aren't really keen on you letting mixing and matching their data with another.
So I think there's, there's a lot of work to be done on you. more granular understanding of omni channel purchasing behavior. Now, some retailers have tried to get around that. Amazon's created their own panel and they're basically reimbursing a couple hundred thousand of their shoppers for scanning their register receipt.
There's a lot of, but there's a lot more work to be done there to have a better understanding of omni channel. I also don't see linkage to data sets outside of retail. Like, the blending, what happened during the pandemic is the blending between food service and QSR, Uber Eats, DoorDash. I want to understand that.
Thank you. What happens to my in store purchasing when, when all of a sudden I see an uptick in, in food [00:47:00] service orders? Those are the kind of things that I think are going to more richly deliver information and help brands make a bigger decision. The other one, the other area I'm really interested in, a lot of Advertising platforms are talking about it.
Publishers are talking about it. It hasn't been widely embraced. It's this concept of non endemic advertising. That means brands that don't sell on a platform advertise on it. Best example, I go to Amazon to advertise my cruise line. You don't buy cruise, cruise line tickets or adventures on Amazon, but the audience that they offer and the ability to target an audience, highly desirable.
So how do I go and do all of that in measuring?
Sean Cline: And it's a hell of a lot of traffic too, right? There's a lot of people going to Amazon.
Peter V.S. Bond: Exactly. So that to me is another unexplored data set. for, uh, for trying to [00:48:00] find audiences that are meaningful to you as an advertiser, but I don't necessarily sell through that platform.
I just want to talk to them. And now, and how do I make the data help me build those audiences so that I can talk to them and, and actually measure performance against them?
Ian Cook: That sounds like an amazing opportunity. Well, Peter, thank you so much for this conversation. It has been as enlightening as it always is listening to you and the rest of the CPG guys.
Thank you for spending some time with us. Thank you to Sean for being here as well. Where else can we see, find, uh, interact
Peter V.S. Bond: with you, Pete? Yeah, so I'm definitely on LinkedIn. Uh, you can just look for Peter V. S. Vaughn. Uh, certainly go to the CPG guys, cpgguys. com. You can subscribe to it on Any major or minor platform, you can ask your virtual assistant at home, Hey Alexa, play the CPT Guys podcast, and she will.
So we're, we're just about, we're [00:49:00] ubiquitous. Uh, you'll see our swag everywhere, as for those of you watching on video, you, you can see it here, but um, those Those are primarily the major mechanisms and then, you know, you see me at a trade show, come up and say hi, I love that. Maybe a TSA line, not at 5. 30 in the morning because that's really not at my best, but I often get stopped on flights, particularly to Benville, Arkansas, surprise, surprise.
But yeah, anywhere, anything like that is great.
Ian Cook: Fantastic. I recently made my first trip to Bentonville. Did not have the chance to see you, but hopefully I will on a future flight. Again, thank you so much. Thank you, Sean. It's been an absolute pleasure.
Peter V.S. Bond: It was great. Thank you so much for having me, guys.
Thank you, Peter.
Ian Cook: This podcast is proudly sponsored by SEEK, the leader in cloud based creation and delivery of industry focused insights. Thank you for listening. If you liked it, Please feel [00:50:00] free to rate and subscribe wherever you get your podcasts.