Latest / Retail Media Breakfast Club / AI Is Stealing the Browse: How Retail Media Wins the Data War in the Age of ChatGPT
Transcript
- When AI claims browsing behavior, commerce media must compete on its data advantage === [00:00:00] Kiri Masters: A year ago, roughly 7% of website referrals to US retailer sites came from chat GBT. By October of last year, that figure had climbed to 16% more than double according [00:00:15] to SimilarWeb. And it's not even just online shopping that's changing. Acosta groups, shopper insights found that 18% of its shopper panel used LLMs like chat, GBT, Gemini, and [00:00:30] Perplexity while shopping in store over the Thanksgiving weekend last year. [00:00:35] Now I've written and talked a lot about how AI enabled shopping poses an existential challenge for retailers across multiple fronts when it [00:00:45] comes to their media revenue. But I'm not even talking about full throttle, agentic shopping, where a bot completes the transactions or reorders. Your weekly essentials consumer behavior has already [00:01:00] changed In that the early stages of many shopping missions now happen inside an LLM. I'm ~gonna share a personal story about what that looked like for me recently. I, I'm gonna share a personal story of how that looked for me recently as a way to explain, I'm gonna share a personal story of what that looked like for me recently, and then also.~ [00:01:07] ~I am ~gonna share a personal story of what that looked like for me recently, but then we're gonna talk about how retailers and commerce media [00:01:15] networks can compete in this new cosmic reset of retail media. Let's jump in. [00:01:21] [00:01:22] Kiri Masters: A month ago, I was in a rut, a hair rut. Have you ever been in a hair rut? I was going through the [00:01:30] motions, doing the same thing over and over 10 years, nothing had changed, and then my hair dryer exploded. This is a sign. I am going to shake things up. I was so [00:01:45] excited I'd heard about new technology like the Dyson air wrap. [00:01:50] Well, awesome. I'm not going to drop $800 on a new hairstyling tool, so I went to Chatt BT to brainstorm [00:02:00] and figure out what's new, what's exciting, what's gonna get me out of this hair rot. I give chat GPT my criteria. I want it to be under $200. I want it to not be super high heat, and it needs to be beginner friendly [00:02:14] Chat [00:02:15] helps me narrow down to three options, then two options, then one option, and we land on. The shark glossy. I go to Best buy.com. Which was one of the retailers that chat, GPT, suggested to me [00:02:30] straight from chat. I check out [00:02:34] and my hair has never looked better. So what's the problem here? Best Buy. Still made the sale. Sharks still made the [00:02:45] sale. And I went through this entire purchase journey faster than ever. [00:02:51] Well, the problem is that I never browsed Best buy.com. I never did any comparisons. I never used any keywords and [00:03:00] didn't do any of my research on a retailer website. That's important because all of that activity is an essential signal that ties into a retailer's first party data set that they then use [00:03:15] their offsite retail media ex audience extension capability, [00:03:20] and with their onsite advertising to help build relevant audiences for advertisers. And thirdly, since I'm not [00:03:30] browsing across all of these pages, I'm really on two or three different pages, the entire journey, the PDP, the checkout page, and maybe an order confirmation page. That's it. So I'm not seeing [00:03:45] much, if any, sponsored product ads on site. [00:03:48] So that is the challenge, this AI enabled shopping journey, which I'm sure listening to my shark glossy story, it's resonating with more than a few of you. There are fewer [00:04:00] signals, there are fewer surfaces, and there is a fuzzier value proposition for brands that are already disenchanted with retail media measurement. [00:04:08] Now I wanna be clear, this isn't a funeral for retail media, but it does demand a different playbook [00:04:15] from existing commerce media players. One where they double down on the data moats that still hold. [00:04:23] So let's talk about four ways that commerce media can thrive as shopper [00:04:30] behavior fragments and condensers. Number one, fuel the inspiration stage. The AI informed journey introduces a formidable intermediary at the research and consideration [00:04:45] stages, but the inspiration still starts elsewhere. I didn't wake up one day craving a fancy hair tool, [00:04:54] social media, word of mouth, and brand advertising for the Dyson Air Wrap. [00:05:00] First put that idea in my head that there was something better out there, and that's where offsite commerce media works, retailers use loyalty and transaction signals to find in-market shoppers and [00:05:15] brand switches and other types of audiences and brands activate against those audiences on offsite channels [00:05:22] like connected tv, social and search. And the retailer then closes the loop with outcomes that media buyers really care [00:05:30] about, like incremental sales neuter, brandand sales, and share lift. Even if the shopper detours to an LLM in their purchase journey for a gut check like that Acosta group data that I mentioned before where people are pulling [00:05:45] out their phones in store and checking with LLMs before they make a purchase, this model still holds up because the conversion still lands in the retailer's lane. [00:05:56] The receipt is the receipt. The only way retailers [00:06:00] blow this up is by killing the golden goose. If everyone can buy the same audiences everywhere, or the take rates on this type of media, get out of control, media buyers will feel burned. Exclusivity of [00:06:15] signals, surfaces, or outcomes is the only way to win here. [00:06:20] Bain and Company. Gave some advice to retailers here that I think hits the nail on the head. Retain ownership of data, retain ownership of fulfillment, and the [00:06:30] checkout. If you control the checkout, you control the receipt, and the receipt is the scoreboard. Retailers know that a marketplace model can dramatically boost [00:06:45] product assortment, shopper engagement, and total revenue. But to get the most out of your marketplace, you need an ad tech solution that can really engage sellers. Miracle Ads is powering the future of retail [00:07:00] media for leading retailers to activate both three P Sellers and one P brands. [00:07:06] Learn more@miracle.com. That's M-I-R-A-K l.com. [00:07:13] Number two, [00:07:15] data collaboration across commerce media players. If shoppers stop leaving as many clues on retailer sites, commerce media networks have two options except. Less signals coming in or [00:07:30] bring in other data that adds context. Some of that context might come from other retailers. [00:07:37] Even the largest network in a category would benefit from. Data collaboration is in a clean room environment where one [00:07:45] retailer might see replenishment patterns. Another might see high intent category behavior. Together they can prove whether advertising across both drives more new to brand buyers, or if they're reaching different people, or whether exposure in [00:08:00] one place lifts sales in another. [00:08:02] And this is how smaller networks can act big without merging or losing their competitive advantage. Other context might come from non-retail commerce networks, [00:08:15] payments, banks, travel experiences, accommodation delivery, other purchase adjacent environments that observe real economic and behavioral data. [00:08:27] These partners can help fill in missing [00:08:30] intense signals that are upstream. [00:08:32] I am gonna move through the next two pretty quickly 'cause we're running low on time. If you want to get a little bit more detail on these, check out the full blog article that accompanies today's podcast. So number three is Product Truth [00:08:45] as a media lever. [00:08:46] When discovery happens offsite and shoppers narrow down to ask skew before they even hit a retailer's site, the old retail media playbook of keyword targeting and browse driven sponsored products, that all [00:09:00] shrinks. So what's replacing it is answering a simple question. Who can reliably fulfill what the shopper wants right now with the least risk? [00:09:10] And that is what we can call product truth. It's the [00:09:15] in stock status, accurate price delivery, ETA pickup windows return policy substitutions. These become important in an age of perfect competition, that LLMs offer. I know everyone's prices. I know [00:09:30] everyone's delivery policies. The reason I picked Best Buy in my sharp, glossy example is because I trust Best Buy. [00:09:37] And, you know, there wasn't any price competition between retailers. So I picked Best Buy out of [00:09:45] a very complex, um, mental arithmetic that I, I did. [00:09:50] But to be clear, retailers can compete on many dimensions beyond unit price. It could be service, it could be picking up my old unit [00:10:00] and recycling it responsibly. It could be, it could be that I can just physically pick it up from the store that same day while rather than waiting for it. [00:10:08] Product truth is critical for advertisers because advertisers, only wanna serve an ad when the [00:10:15] item is in stock and available and and knowing that the product information in these LLMs and on the, on the product cards is completely accurate, so this is an area where retailers play a critical role in [00:10:30] maintaining accurate product. Data feeds and information to ensure everything is still accurate in this new buying context. Number four, in store as the durable arena. Here's the twist. In an AI [00:10:45] shaped shopping world, in store is really the most defensible part of the commerce media stack. [00:10:51] No LLM can replicate physical presence, sensory context, or real time proximity to products. When everything else becomes [00:11:00] abstract and chronically online, the physical store becomes more valuable, not less. [00:11:06] Now what? Many retailers are embracing this new shopping paradigm and building for an agentic future, but it does [00:11:15] call into question the future of media businesses built largely on onsite sponsored product ads. As of today, in the US 70% of all retail media ad spend goes to onsite [00:11:30] retail media. [00:11:31] The commerce media networks that thrive in this reset will be the ones that recognize what they still uniquely own. Transaction data, identity, loyalty signals, physical stores. That ownership [00:11:45] demands different muscles than the sponsored product playbook that built retail media's first decade. [00:11:51] As AI enabled shopping claims, the early and mid funnel commerce media needs to think about a new [00:12:00] playbook that is grounded in its inherent data advantage. Thanks for listening, and I'll catch you tomorrow. [00:12:06]