Subscribe
Share
Share
Embed
10 minutes of expert insights every weekday. Your morning ritual for staying ahead in retail media.
Kroger On Why Context Is King
===
[00:00:00] Kiri VO: A few weeks ago, Kroger Precision Marketing announced a partnership with Google that lets brands activate Kroger audiences inside of YouTube and DV360, with measurement flowing back through a conversion API that is tied [00:00:15] to Kroger sales. On the surface, it's a media story, another off-site channel, another way to close the loop on awareness spend.
[00:00:23] But scratching below the surface, it's really a story about context, [00:00:30] a system that is anchored in two decades of loyalty data that knows what shoppers actually buy across both digital and physical aisles. And that is the piece of the stack that Google and other infrastructure [00:00:45] players can't build themselves.
[00:00:47]
[00:00:48] Kiri VO: And that's the under-discussed shift happening in commerce media right now. There's a lot of capabilities out there that are starting to look the same, but what retailers uniquely [00:01:00] own and what the more sophisticated ones are building around is context. Context is not just data. It's an understanding of who the shopper is, which is stitched across devices, channels, and household members. [00:01:15] It is what they actually buy, not just what they click on. It is where they buy it, online, in-store, click and collect, and finally, how that behavior changes over time across years of purchases, not just a single session.
[00:01:29] [00:01:30] Models can be rented. Identity vendors can be rented. The relationship between a shopper and their transaction history stitched together reliably over time, that has to live with the retailer, and that is the durable [00:01:45] asset that RMNs
[00:01:47] can continue to build on. Last week, I sat down with Christine Foster, who is the Group Vice President, Commercial Strategy and Operations at Kroger Precision Marketing, and Jon Flugstad, the Chief [00:02:00] Commercial Officer at MetaRouter, which is the data infrastructure partner powering that integration, and we talked through what they actually built and why it matters.
[00:02:10] The full conversation is available to replay if you wanna listen to that or watch it, [00:02:15] but I'm gonna share a few highlights from this conversation around shopper context specifically. Let's go.
[00:02:22] Christine: For our, for, from my point of view and from Kroger's point of view, you know, we're building on the foundation. Kind of goes back to the foundational point [00:02:30] of, um, the data here. The, it's built on a shopper loyalty program that's ex- existed- Mm-hmm
[00:02:36] for more than 20 years. Mm-hmm. 95% of all Kroger transactions are connected to that loyalty account. [00:02:45] That means that this data is, you know, verifiable, if you will, and there's a lot of behavioral, um, kind of i- inferences here that are really critical to marketers to understand, [00:03:00] um, about their audiences and about their customers who are our customers in a lot of cases.
[00:03:06] Um, I think most retailers, when you, when you talk about, like, being able to just copy and paste this, most retailers [00:03:15] cannot, or don't have that level of confidence in their data sets, not saying that they're bad. But I, I think that for us gives us a lot more, um, confidence in kind of who we are and, and why, uh, this was [00:03:30] a good first step for even YouTube to take with, with us.
[00:03:34] Um, and, and with, you know, 80-plus percent of US retail sales still happening in the store, that piece is a piece that [00:03:45] transparently digital marketing has, has missed. Mm-hmm. And so you're, you're able to connect this online to offline world through this data set and through this, um, connection into YouTube and Google in a way that [00:04:00] just hasn't been possible in the past for marketers.
[00:04:03] And so I think when we think about that, that is really an important piece for everyone to keep in mind as they're thinking about doing similar things.
[00:04:13] Kiri: John?
[00:04:14] Jon: Everything [00:04:15] Christine said, and just sort of like building the statue over a year with, like, globs of clay has, like, added into, like, something that isn't, like, necessarily coherent a lot of the time from a technical foundations perspective, and some of those were choices that were required.
[00:04:29] I'm not talking, [00:04:30] not talking about Kroger specifically, but in general across retail media, it's like-
[00:04:32] Kiri: Right ...
[00:04:33] Jon: I have this platform for onsite sponsored products. I have this platform for onsite display. I use all these different platforms for offsite activation. I have this platform for collecting data. I have an identity partner that I use for onboarding.
[00:04:44] I have another [00:04:45] identity partner I use for, like, household identification. There's all these different components that don't speak to each other, and so the reconciliation work of trying to make your data work, let alone make it work in, like, really low latency for the type of sharing you wanna use for targeting [00:05:00] and the algorithms we were talking about earlier on, like, outcomes It's just hard.
[00:05:03] Like, you're trying to, like, operate on an environment that doesn't wanna be reconciled. And so that's where some of the architectures are, are, are changing, is, okay, well, you know, a [00:05:15] layer can connect all these things and make sure that the pipes work, both externalized destinations, internal destinations, putting data out, pulling data in, a lot...
[00:05:23] And making sure consent is honored, aligning it to identity, collapsing anonymous and known identities in such a way that you have [00:05:30] this foundational data layer that can feed context anywhere, and that's the, like, y- what, what's, you know, the, the token comment on AI, like you have to have it in something.
[00:05:39] But, like, going forward, that's the retailer's biggest strength. Like, they have context over what the [00:05:45] shopper does and who they are that no one else has, that they can then control across any of those channels and, you know, frankly, replicate. And control over that is so important. But, you know, models will be out there in AI, there'll be, you know, probably commoditization of the various [00:06:00] models.
[00:06:00] Context is unreplicable. It's the thing that, that the retailer owns that they're now, like, orienting the architecture around because it's the gold. And so the fact that that's changing in a fundamental way and moving from this disparate set of [00:06:15] systems where some of it's owned by others or rented by others or, you know, outsourced.
[00:06:19] It's like, no, no, no, that's mine. Yeah. I'm the retailer, I own that. And so I think that's one of the limitations from moving overnight to this new kind of, you know, model that Christine's talking about. Just a lot of [00:06:30] debt that's piled up and, you know- Yeah ... fundamental trades to be made over control, Did you know that leading retail media networks drive 85% of their [00:06:45] ads through mid and long tail advertisers?
[00:06:48] Kiri Masters: Miracle Ads provides full funnel ad formats tailored to both one P and three P advertisers leveraging unique AI capabilities that provide unprecedented [00:07:00] levels of relevance and engagement. Retailers who want to capture ad spend from the long tail of three P Marketplace sellers use miracle ads in their tech stack.
[00:07:11] Learn more@miracle.com. That's [00:07:15] M-I-R-A-K l.com.
[00:07:18]
[00:07:18] Kiri VO: And this is where Christine and John make a point that I think a lot of RMNs need to hear, because the way that Kroger activated this with Google wasn't through a portal that Kroger [00:07:30] built. It was self-serve straight into the accounts where brand teams are already working
[00:07:35] Jon: So it's a pretty exciting time for commerce media. But I guess to actually answer your question on the technical components, um, have you ever seen that meme of like that really tall [00:07:45] deck where it has like those like very flimsy tall beams like holding it up?
[00:07:48] It's like this like about to tip over like large deck. It like the, like the, the point of it is there's multiple ways to build it. There's like a right way to build it and a wrong way to build it that like supports your act- like what, what you actually wanna do. And [00:08:00] so I think instead of like architectural details or what needed to be true, like what were the kind of guiding principles of building it is useful.
[00:08:07] Like, there are ways you can do it and accomplish similar outcomes that aren't as durable or maybe as strategic. But what I think was true in the case [00:08:15] of KPM was they set out to design it in a way that gave them control. So with g- you know, we are deployed within their private cloud. Um, and so they have maximal control over data and what's used for any kind of like matching.
[00:08:29] [00:08:30] Second, the thing that Christine mentioned is near real time, so you know, these signals matter. They're training the outcomes that these, that the algorithms can drive. And so, you know, having less lag on how that data is shared and flexibility over how it's routed, even down to brand [00:08:45] level accounts for self-serve, is really a kind of a design principle here that makes it really flexible and, and pretty innovative.
[00:08:50] The third is representative of kind of all sales, not just online, which, you know, for a grocer obviously, for a long period of time, like there was, you know, optimization around the online [00:09:00] sales, but in-store remained the lion's share of overall sales. So how are you matching in-store sales in such a way that allows you to drive the outcomes and measure it?
[00:09:08] And then the last thing is just building it around a way where you have operational and financial leverage. You know, there [00:09:15] are partners or others in the industry that, you know, dare I say, charge you serious fees around kind of these similar types of capabilities. And you know, if you can enable it in a way that drives leverage as you grow this thing, like those are kind of the, I think the overarching design principles of how it was approached, and [00:09:30] I think it's the r- the right way to do it.
[00:09:31] Beyond sort of the technical nitty-gritty, it's like more like design principles of how you approach it technically overall.
[00:09:37] Christine: Uh, I wanna go back to your omni-channel kind of point and how powerful that visibility is for brands, John, [00:09:45] and, and why we are partnered with you on making that happen.
[00:09:49] Because to your point, I'm, I am not an online-only shopper. I would guarantee none of us are. Um, and there's probably a balance, um, for most [00:10:00] shoppers, right? But a lot of transactions are still happening in store. And to John's point, um, having that view power what you need to do with your, your strategies and how you optimize your [00:10:15] campaigns is, I think, just the next step in, in our evolution as an industry and, and the way we should be moving.
[00:10:22] Which is to say that full view is going to help us all make better decisions and, and optimize [00:10:30] campaigns and marketing spend in ways that we couldn't before. So going kind of two prong, right, going back to your question to me, Kiri, is like what can people, you know, marketers do that they couldn't do before?
[00:10:39] But also, that may not have been possible before 'cause we didn't have partners like Meta or Audor [00:10:45] before, right? Like, there is a, there's a lot of advancing happening technologically that is important for all of us to, to get there, and I'm excited, um, on kind of where we'll go from here.
[00:10:56]
[00:10:58] Kiri VO: So this new announcement could be [00:11:00] read as a media story, and it is that, but the more interesting read is what it implies about where retail media is heading structurally. What this conversation makes clear is that the real differentiation isn't in the formats or the [00:11:15] channels at all.
[00:11:16] It's in the depth, cleanliness, and architectural accessibility of the context layer underneath.
[00:11:23] The retailers who win in the next phase of commerce media aren't gonna be the ones with the most [00:11:30] ad products. They're gonna be the ones whose context can flow into YouTube, into DSPs, into agentic surfaces, into wherever shoppers are actually making decisions without losing fidelity along the way.
[00:11:44] And that [00:11:45] is an infrastructure problem rather than a media problem, and it's one of the few problems in this industry where being early actually compounds. Thank you to Christine from Kroger Precision Marketing for joining me and Jon [00:12:00] Flugstad on that live. If you want to listen to the full live stream, we'll link up to it in the show notes, and thank you to MetaRouter for sponsoring that event.
[00:12:10]