Subscribe
Share
Share
Embed
10 minutes of expert insights every weekday. Your morning ritual for staying ahead in retail media.
Edited snippets with comnetary
===
[00:00:00] Kiri: The conventional wisdom says that consumers will hate ads in their AI assistance. That inserting commercial content into chat, GPT or Gemini will erode the trust that makes these tools so [00:00:15] valuable. I am not so sure on that. I recently joined Mike Shields on the next in Media podcast to talk about ads coming to LLMs, and I made an argument that surprised even me, Not only could [00:00:30] these ads work, they might actually be better than what we've ever seen before. But here's the thing, if I'm right about that, it kind of creates a problem for retail media that almost nobody is talking [00:00:45] about.
[00:00:45]
[00:00:46] Kiri: Let's listen into part of my conversation with Mike.
[00:00:50] Mike: Is it, um, do you think it's worth questioning whether people will respond to ads the same way and whether you're gonna have the same just breadth of advertising in a, in, in [00:01:00] experiences that are about sort of like, they're, they're different than, I want 20, 20 blue links.
[00:01:06] I wanna have an answer. Like, are we, are we sure this is gonna be as powerful as we think?
[00:01:10] Kiri: Great question, and yes, the, the proof is gonna be in the pudding. It all [00:01:15] depends what these ads actually look like, right? As, as to whether people will engage with it, if they will still feel like their AI assistant is a trusted collaborator rather than someone just, you know, sleazy salesperson trying to pitch things to [00:01:30] them.
[00:01:30] So I think it will really depend on what that AD U unit looks like, how it is served. I personally am very optimistic that we don't necessarily need, they won't necessarily look like the sponsored products [00:01:45] of the past for a few different reasons. One, with AI, not so much now, but in, in the future, I believe we'll, we'll have much more multimodal engaging.
[00:01:56] Surfaces to interact with ai, voice, [00:02:00] video, things like that, try ons.
[00:02:02] Mike: Right? We shouldn't be limited our, we shouldn't limit our thinking to what we see right now.
[00:02:05] Kiri: Exactly. So I, and I think, you know, for certain types of shopping behavior, having that sort of multimodal ad would be really helpful. Like, for [00:02:15] example, um, I would love to be told about a new product drop from my favorite clothing brands and have it, you know, shown on my, uh, you know.
[00:02:25] To show me what it looks like and just have a popup like, Hey, Kiri, one of your favorite [00:02:30] apparel brands have just dropped a new line. Right. So that would actually be cool. Another thing is around context and, um, how much these LLMs know about us and ads are [00:02:45] annoying when they're irrelevant. Right, right. A lot of people love the ads on.
[00:02:49] Instagram for example, because they're super relevant. I'm like, I didn't know that I ever wanted that thing. Yeah. And now I do. And, and so it can feel additive when it is [00:03:00] super relevant. And you think about all of the context that these LLMs have, we're sharing not just our, you know, requests for recipe ideas, but we're sharing our hope streams fears.
[00:03:12] Marital counseling [00:03:15] questions.
[00:03:15] Mike: Right, right. Literally. Literally having therapy sessions. Yeah,
[00:03:17] Kiri: exactly. So I think that there is, I am an optimist. I think that ads in LLMs have the potential to actually be really cool and big question is [00:03:30] if they sort of take that direction and run with it or you know, try and get some, try and sort of copy what has been done on other platforms.
[00:03:40] Mike: That's a really interesting point. I hadn't thought of it that way, but because it is, um, there [00:03:45] is a trust, there is a unique level of trust. I mean, again, it's really early in our journey here, but like, people are really personal and they be, you know, they sort of buy into their, these products, talking to them in a way that.
[00:03:58] Where traditional [00:04:00] search you could be. And it took a little while before Google became like nakedly commercial, but then I, I don't think people had that expectation that it wouldn't be where they'll have to be careful how hard or how fast they go, [00:04:15]
[00:04:16] Kiri Masters: Miracle Ads is the only retail media solution designed for both one P and three P Marketplace brands. Why does that matter? Marketplace sellers demand a seamless advertiser [00:04:30] experience that still offers full funnel ad formats, and retailers need a flexible solution that allows you to scale your media business.
[00:04:39] Learn more@miracle.com. That's M-I-R-A-K [00:04:45] l.com.
[00:04:48] Kiri: So that's the optimistic case. LLMs have contexts that no other advertising platform has ever had. They know what we are worried about, what we're planning, what we're searching for across every [00:05:00] category of our lives. But later on in the conversation with Mike, I raised something that's been. On my mind, if LLMs accumulate all of this rich intent data and they figure out how to [00:05:15] monetize it, what does that mean for retailers who've been building their offsite media businesses on exactly that same data?
[00:05:24] Amazon was early to the game. They've trained brands on what to expect. And other retailers [00:05:30] just, you know, they, they can't compete with. What Amazon has done specifically, so there's other, the, the two other big areas of retail media, which are not nearly as big as onsite sponsored product ads are offsite retail media, which is, you know, [00:05:45] activating a retailer audience on CTV or social or other, other platforms.
[00:05:50] And finally re uh, in-store retail media, which is still very, very small, but people are very excited about it. That's very resilient to AI enabled shopping, by [00:06:00] the way. Right? Right. So I think there's a good reason to be excited about that. But actually, I wanna run this by you, Mike, because I, I, I wanna get your opinion on this.
[00:06:06] Sure. When I think about offsite retail media. For, um, that, uh, it's not as high profit margin as, [00:06:15] as onsite, but in the event that people are doing more like actual transactions through LLMs, I see that as being, uh, as creating a big risk for retailers offsite media. Now another [00:06:30] party has that transaction data and they have that intent data and the the
[00:06:36] do
[00:06:38] Mike: that's I. In general, I, when people talk about off platform, you're right. It's, they, they want, [00:06:45] they wanna, let's use that data all over the place. But I, my mind always go, goes towards the open web because that's, um,
[00:06:51] Kiri: yep.
[00:06:52] Mike: That, that's where you think that's, there's the infrastructure's, right, that is already there.
[00:06:55] And it, and it's, the open web is hurting theoretically. Right? Like the so, but, [00:07:00] but in my experience, at least from, I don't see retail media ads all over the place. Like, I think, like I would expect maybe now that could just be me. You wouldn't know it. Would I know it? I guess, yeah, I guess. I guess I'm not sure if I would feel it or not, but, [00:07:15] and in CT.
[00:07:22] I I, I do think that if, if, uh, shopping takes off through chat experiences, that that can [00:07:30] hit retail media in general. But I hadn't thought about. I I, but I I think you're probably right. You were, you were gonna, the brands will go to what works and what is closer to a transaction than something that feels a little bit disconnected so that, that would make it vulnerable.
[00:07:42] That does make sense to me.
[00:07:43] Kiri: I think it would, it, [00:07:45] it might bring the premium. Uh, CPM down a little bit. I, I think because there'll be a bit, a bit of competition for that, um, that audience data. So, you know, now, Chacha Pinos, I'm an Atlanta, I'm a [00:08:00] mom. I live in the suburbs. I kind of shop this stuff. I've been right, searching for new recipes.
[00:08:04] I've been searching for, you know, high protein lunchbox, snacks or whatever. Um, Walmart also has that search. Data. Data. And so they [00:08:15] could use, you know, uh, uh, a, a protein brand could retarget me on Disney with a snack commercial or something like that.
[00:08:25] Mike: Right?
[00:08:25] Kiri: But also Chatt knows that. And so I, I think this all comes down to if [00:08:30] the LLMs choose to pursue an offsite, um, media.
[00:08:34] Business. To me it makes sense because then you're not sort of quote unquote cluttering the chat experience with ads.
[00:08:43] Mike: Right,
[00:08:43] Kiri: right.
[00:08:44] Mike: You, you'll have a lot [00:08:45] more surface area with, with people. Were theoretically more receptive that That's actually really interesting. I hadn't thought about that.
[00:08:50] Kiri: We haven't seen any LLMs talk about it.
[00:08:54] Certainly not release it. T has even said something very specific like, we will never [00:09:00] sell consumer data, uh, to advertisers. Yeah. I'm like, there's a lot of never in this business that don't pulls. Yes, well true. Um, to me it just seems like if, if I was in charge of monetization at one of these lms, it's a path I would seriously consider, but maybe [00:09:15] there's a good reason why they haven't.
[00:09:16]
[00:09:17] Kiri Masters: Retailers have spent years building the pipes to activate their shopper data across CTV, social and the open web. That offsite business was supposed to be the next growth engine as onsite inventory [00:09:30] gets saturated. But if chat, GBT knows that I'm a mom in suburban Atlanta searching for high protein lunchbox snacks, and they decide to build an offsite advertising business instead of cluttering [00:09:45] the chat, suddenly retailers aren't the only game in town.
[00:09:49] With that data. I'm not saying this will for sure happen. I don't have any inside track on this, but if I were running monetization at an LLM, it's [00:10:00] a path I would seriously consider. I really enjoyed this conversation with Mike Shields from Next In Media. I personally love his substack and podcast. I think you should definitely check it out.
[00:10:12] If you like this one, we'll link up to [00:10:15] it in the show notes. Thanks for tuning in and I'll catch you tomorrow.
[00:10:18]