Retailgentic, presented by the ReFiBuy Media Network and hosted by Scot Wingo, is where Agentic Commerce meets retail innovation. We help retailers and brands prepare for the future of agent-driven shopping with actionable news, expert insights, dxis, forward-looking predictions, and practical guidance on Agentic Commerce Optimization (ACO).
I hope so. I hope I hope we see some new and creative ad formats. I also hope we see some new and creative targeting options as well. Because, you know, the current the current MO is I'm using all of your browsing history and information about what you clicked on what you might be in market for, rather than the context of the chat itself and the content that I'm looking at right now.
Speaker 2:Welcome to the Retailgentic podcast, where we explore the intersection of retail, ecommerce, and AI agents, which we call Retailgentic. Our mission is simple. We are shining a bright light on the new Agentic AI shopping megatrend, which will disrupt the trillion dollar digital retail market. The stakes are high. Google, OpenAI, Microsoft, Perplexity, Amazon, Apple, Meta, and Anthropic are all fighting to be the consumer AI shopping agent winner.
Speaker 2:There are gonna be some tough decisions for online brands and retailers to make. What happens to checkout? How should payments work? Could you continue to invest in retail media networks? What does the world look like when over half of your traffic is not from humans but from AI agents?
Speaker 2:These existential questions are the tip of the iceberg. On this podcast, we're talking to industry experts, analysts, and influencers to get answers to these questions, identify trends, and best practices. This is Retailgentic, your podcast partner in navigating our AI shopping agent future. And now here's your host, Wingo.
Speaker 3:Welcome to this episode of Retailgentic. On this episode, I welcomed back Kiri Masters. Kiri is my go to on all things advertising as they intersect the world of e commerce. She is a well known expert on... She had an agency, so she knows...
Speaker 3:She's kinda lived the day to day, and then she's an expert on retail media networks. In this episode, I I wanted to talk about the topic that is top of mind with most brands and retailers I talk to. They've gotten their head around Agentic Commerce, and now they're all thinking about what's next. And the logical conclusion everyone talks to is we're seeing the worlds of traditional advertising, retail media networks, which is the kind of product retail flavor of that colliding with Agentic Commerce. What is that gonna look like?
Speaker 3:What are some of the things out there? And Kiri was kind enough to come on and share, first, an update of what's going on in Retail Media Network. I thought that would inform the decision, so we get a brief update on that. Also, if you recall, I had a prediction that my fourth prediction out of 11, number four, was that retail media networks will feel the eugenctic commerce pain. So I kind of wanted to check-in on that and see if that's happening, and you'll see where Kiri lands on that.
Speaker 3:It's still early in February, so we still have time if they're not feeling it. And then I go through I've been keeping a very close catalog. And this does I know a lot of you listen on audio, which is great. This does have a visual element. It will be available in the sub stack as a PDF, you can check it out later.
Speaker 3:And we... I very keenly always try to kind of at least describe what we're looking at so you can you can imagine it if you're just listening as you're going about your day. So we go through that, and I catalog all those things that have been announced by Amazon, Google, ChatchBT. And then I pick Kiri's brain on what is this gonna look like as we kinda smash the best things that brands have learned from the retail media network world and what we know that the the Agentic Commerce engines are doing. How do we how do we marry all that together, and what is this future of these two worlds colliding gonna look like?
Speaker 3:I think this is probably your best view into the future that you'll find anywhere. And I really enjoyed... It kind of is a little workshoppy. It's a little bit of a different format, but I think you'll enjoy hearing us riff off each other, hearing where Kiri and I disagree and agree. I I think that this is the kind of discussion that is happening, throughout the industry that a lot of times is is kind of, you know, in the back hallways of of trade shows.
Speaker 3:So, Kiri and I are super transparent and are able to just hop on and do this type of thing, which is, you know, not what most people at trade shows are doing. So, I hope you enjoy kind of how this conversation unfolds. I learned a ton, and I'm sure you will as well. And now without further ado, my conversation with Kiri Masters about what's going on with the intersection of ads and agentic commerce. Enjoy.
Speaker 3:Welcome to this episode of Retailgentic. I'm your host, Wingo, and I am joined by one of our favorite guests, Kiri Masters, the host of Retail Media Breakfast Club. Hey, Kiri.
Speaker 1:Hey, Scot. Great to be back.
Speaker 3:So I was looking and we haven't chatted since mid October, which in AI era is a decade. So, man Gosh. How's the last ten years been for you?
Speaker 1:Yeah. Just a dream. Just a dream. Yep.
Speaker 3:So one of the reasons I wanted to get you on is I am getting a lot of questions around, the intersection of Gen AI, Agentic Commerce, and advertising. And I thought we could start, well, first of all, if folks are new, give give the quick thumbnail of who is Kiri, what's Retail Breakfast, Retail Media Breakfast Club, and where do you spend your time?
Speaker 1:Yeah. So I started Retail Media Breakfast Club a year ago. Every day of the week, I wake up and I try and find something to say about retail media, which is, you'd be surprised. It's actually a very long list of, dramas, squabbles, disagreements, and all kinds of interesting things happening in retail media, which a fascinating industry. Increasingly, I've been sort of talking about the intersection of Agentic Commerce or AI enabled shopping with retail media because I think it presents some unique threats and opportunities for for retail media.
Speaker 1:Maybe a retail media three point o on the horizon with this like changing shopping journey. And so what that looks like is a newsletter and podcast every morning Monday to Thursday at 6AM called Retail Media Breakfast Club.
Speaker 3:I'm intimately familiar with the squabbles because I got pulled into one about I guess, around the time we did our last podcast. The, the... So lips... Lipson and that eMarketerly, Sarah Metzger.
Speaker 1:Sarah Mausanne. Yeah.
Speaker 3:Yeah. Oh, yeah. Her. Yeah. They both came at me, they're like, Agentic Commerce is fake, and it's a hallucination.
Speaker 3:I'm like, woah. Like, why am I the guy you're picking on? So we're actually gonna do a, a battle at Shop Talk about that, so that'll be delightful. Yes. Yeah.
Speaker 3:Battle Royale, the thriller in Manila.
Speaker 1:That'll be... That that'll not be one to miss. I think I'm gonna be doing a battle at that same time as well about retail retail media rather than Agentic Commerce. So I'll be I'll be up there on stage with my my my boxing gloves on as well.
Speaker 3:Yeah. Fun. So I thought we could talk by bringing us up to speed on what's happening in your retail media world. What are what are people in that world thinking about? I know, Andrew thinks it's hallucination, but the people that actually believe Agentic Commerce is real, what are and I kind of famously had Roy Rubin on the show, and he basically said websites won't exist in five years.
Speaker 3:So that's one perspective. And if you're in retail media, that's kind of bad news for you. But I think that's an extreme view of things. What's kind of the... Give us kind of update of what's going on in top of mind over in retail media.
Speaker 1:Yeah. So I think, the way I think about it is there's a couple of there's a couple of ways that Agentic Commerce and Agentic AI enabled commerce threatens the current state of retail media. So let's talk about on-site, off-site, and then in store experiential kind of stuff. So on-site retail media, that's the sponsored product ads that relies on people visiting a website or an app and seeing an ad. And so the threat that AI enabled commerce brings there is if we arrive at a retailer.com or an app later in our shopping journey, after we have already collaborated with ChatGPT and decided what SKU we're gonna buy, we get to a retailer.com later.
Speaker 1:We're not browsing. We're not using keywords. We're not, navigating category pages. There is way less. We're going through way less content on a retailer site before we get to an actual sale.
Speaker 1:So that means less inventory for us to see, less inventory for a retailer to sell.
Speaker 3:So so previously, you had the the homepage, the search engine experience, kind of like, you know, and then like several searches, right? Because people couldn't find anything in e commerce in the olden days. Then ultimately you land on a PDP then, or even then you're like, so I guess SERPs, there's some stuff in there that's part of the search thing and then a PDP. And also in these page views, there's chances for like a hero banner at the top and some stuff in the gutter and... All kinds of things.
Speaker 1:Exactly. All these different forms of on-site ads.
Speaker 3:Do know what percentage that is versus the PDP? Like, is it...
Speaker 1:Yeah. So overall in The US, and this this is different in other markets where there's lots of different mechanisms at play. But in The US, on-site ads make up about 70% of all retail media spending. So this is the big bucket, and this is very profitable for a retailer to run. I mean, they need the technology.
Speaker 1:They might build it. They might buy it. But that might be 70, 80% profit margins for these sponsored product ads. So this is a this is a big contributor to a retailer's bottom line.
Speaker 3:How about like of those SERPs banners? Does... Has anyone split all that out inside that 70% bucket or or?
Speaker 1:That's a little harder to split out.
Speaker 3:Yeah. Okay.
Speaker 1:Yeah. That kind of goes in into that performance marketing budget over overall. I guess you'd have different types of campaigns that you're running. But that that's the threat to that type of retail media. The other threat it is emerging, but I'm sure we're gonna spend a lot more time on this in in this episode, is that there's these new surfaces.
Speaker 1:The LLMs that are now offering their own ad space and that might end up directly competing for that performance marketing budget that brands are currently spending on a retailer.com on-site ad. So those are two threats to on-site retail media. Off-site retail media, this is a little bit smaller, in terms of total ad spend in in The US. I wanna say 10, what are we at? 15%, maybe.
Speaker 1:Lower profit margin on this one, maybe about 40% profit margin.
Speaker 3:And this is where I go to a retailer and I look at a product and now I go somewhere else and I get retargeted that thing?
Speaker 1:Yep. Can be can be like that. Yeah. That would that would be an off-site targeting, but it it... It's using all of the signals that a retailer has, whether that's online or offline to activate on a third party property.
Speaker 1:So that could be on a website. You get those annoying like banner ads showing up for the craft mac and cheese that you just looked at or it could be actually on Disney, watching Disney and Disney has all of these partnerships with retailers and and the advertisers use an, retailers audience data to figure out who to target on Disney for Kraft Mac and Cheese. It can also be on social media, any anywhere off-site where a where a brand wants to know. I wanna target this particular type of consumer and then I also wanna know, did they end up buying that product at that retailer. So that's the that's the closed loop attribution that's very powerful about retail media and that's how it's used off-site as well.
Speaker 1:So the challenge to off-site retail media is, again, we're not going... We're not spending as much time clicking around, comparing, looking at something, researching on a retailer's website. And that's all very valuable information that goes into a retailer's dataset about their audience. So those signals decrease and it means that their off-site audience data pool shrinks a little bit. It becomes a little less enticing.
Speaker 1:They're not getting as much information about who we are and what we're interested in. So that's one challenge to off-site. And then I think the other challenge is these LLMs haven't announced anything yet, but I think they're also out. They're also collecting a lot of information about who we are, what we're interested in, what we're interested in buying. So I think it would be a pretty elegant way for an LLM to solve their monetization issues if they set up an off-site media network audience extension capability.
Speaker 1:Because then they wouldn't need to even put ads in the chat experience. So that's a... The the the thing there that is resilient for retail media is that the retailer is the one that collects the actual transaction data, which, know, the LLMs are starting to get that with, like, a fully agentic scenario. But they're not... That...
Speaker 1:That's to be missing some pretty important pieces, but it is still in in my mind a threat to the current scarcity and premium pricing of that ad type.
Speaker 3:And then your third is in store. Are they at risk or they're hunky dory?
Speaker 1:Hunky dory. Yep. No. This is this is a very interesting space because it's currently a very small percentage of retail media ad spend only about 3% in The US when you take out Amazon because that that sort of sways the numbers quite a lot because they don't really do very much in store at all. So 3% of total ad spend.
Speaker 1:And so there's a lot of, you know, commentary in the retail media space that in store advertising should be much bigger and it it's so effective and why don't we do it more? Well, there's a pretty significant cost to actually setting up all of those screens and things in stores. It takes a long time. You're trying to put all this new technology into stores where people are working and they're trying to restock and there might not be a power outlet where you need it to go. So it's it's complex to do.
Speaker 1:But it it is resilient to to all of the stuff that we're talking about because, yes, people will keep going to stores. You wanna get out and about on a rainy weekend with the kids. You wanna go check out the new kind of apparel trends and that the in store stuff is gonna kind of be completely resilient to this in my mind.
Speaker 3:Yeah, there's been several times when I go to the grocery store and it's always like randomly during the day I'll go and it's me and all the Instacartners. And then I wonder like, how many are they? Are you guys multi, you know, are you subtracting out all the Instacartners seeing your ads? Because they they could care less because they're just like That's looking at that's... Yeah.
Speaker 3:I wonder what percent of eyeballs are are Instacartners that you're you're advertising to kind of ironically.
Speaker 1:That's a great question.
Speaker 3:So so it seems like the real, the biggest threat is the big kind of the honeypot, which is the on-site, because that'll impact the off-site to your point, because you'll have less data. Are people seeing it? Because when I talk to retailers, especially in fashion and beauty, they are definitely seeing a decrease in traffic, period. And they are definitely seeing it kind of, you know, skipping through and going to the PDP. And then all of them, so they realize that there's an elevated experience now that they're having to up their website game to, and they're implementing all these new solutions at kind of what I would call that conversion layer.
Speaker 3:So they're implementing a better search engine. All of them are... Many of them have an agent, but they're not happy with it. And now they can kind of see where Rufus and Sparky are. And they're like, Oh my goodness, we're...
Speaker 3:It feels like several... Most of their on-site agents are really more what I would call customer support, Where's my order? And they're really good at that, but they're not really good at discovery. So I think that could also be really interesting. Another kind of element of this is as they upgrade those things, is that going to also, even when people do come to the site, are they going to be accelerating them to a PDP themselves?
Speaker 1:Yeah, that's interesting. I think there's there's a lot of... It's a great point about it being category specific. Interesting about that being apparel. There's there's some retailers I've spoken with and and they're not really concerned because they don't have a huge on-site business.
Speaker 1:There's some retailers where it's all in store. Think about a discount store, right? Or a TJ Maxx. They they don't get...
Speaker 3:Is Dollar stores.
Speaker 1:Not gonna affect them at all. Right? Dollar stores, etcetera. They're they're not worried and and maybe they shouldn't spend too much time worrying about it. But a a retailer where there's there's a compressed shopping journey that's now happening, I think we'll really need to stop thinking about what does a a a version three look like.
Speaker 1:To be honest, I I think that maybe some of these current revenue streams shrink, but there are more opportunities for brands and retailers to partner together. I've got a whole list. I think it's just going to change. I think, you know, the the the sad thing is for a lot of brands, they have felt for quite some time that a lot of the retail media ad budget that they're spending has been a margin extraction rather than a truly additive place to spend money. And they're kind of forced into it by their retailer partners.
Speaker 1:That is, you know, that it doesn't always happen that way. There's there's some there's some retail media, activations that are truly additive and exciting and a great win win. But there is this kind of reputation that at least some of the time, it's retailers putting on the hard work with brands rather than it being a true win win. And there are win win media opportunities that can be built.
Speaker 3:What do you think about the ad experiment that Chatchippity is doing? So they did their kind of like little warm up kind of like the cats on the roof is like this old joke of how you kind of like ease people into things, that they were going to run some tests. We're seeing them live out there. Several people announced that they're in the program. What's kind of your current thinking on where that's going?
Speaker 1:It looks very rudimentary right now, launching with a CPM model rather than a CPC model, launching with such a limited number of ad partners that they didn't fling open the door and and let everyone in. From what I've what I've heard anecdotally is it's all quite, manual and very much a closed door experiment. That could be because they're just, you know, that they're still getting everything ready and and they're they're building the engine and they're they're building the plane as they're trying to fly it. That's the most sort of credible explanation to me is that they didn't want to wait to perfect the thing. They're sort of learning as they're going.
Speaker 3:One thing I wanted to kind of do a live podcast workshop. This is what I love about Kiri. Was like, Hey, we're just going to come on and wing it. And she's like, Let's do it. And so thanks for that.
Speaker 3:So I'm getting asked by all kinds of retailers and brands to come help them understand, give like, us a position of what's going on in this ad thing. Because a lot of them, once we talk to them about Agentic Commerce Optimization, and especially brands, their next thing is like, is there a way for us to pay for better placement and a different look? And the answer today is no. But then my follow-up is, but I think it's coming soon. Maybe not Chat to BT, but I think we're going have a wide variety of these things.
Speaker 3:And so kind of my framework here is, first of all and if you're on the audio, there's a video, a visual part of this. I'll try to talk through it. So the first part of this is I'm just capturing what we know today. And here I'm showing an image of, when ChatGPT announced that they were going... There's kind of the...
Speaker 3:Don't I know about you, but I've talked about this a thousand times now. I'm kind of burned out on this one. But there's this famous one where they showed this example of someone meal planning, and in there, there is a, the famous Ember Company hot sauce where where they show it in stock in in an ad kind of tied a product ad, product related ad tied to that prompt. Now I do think the big miss here is it didn't have the Buy button on there. But I do think, you know, what we learned in the Amazon era and a lot of my ad thinking is like, What did Amazon innovate?
Speaker 3:And will we see a lot of that come out? And then maybe also Instacart because of Fiji. And I'll talk about that in a second. But I think what will ultimately happen is you want to create an incentive for people to be an instant checkout and on your e commerce platform. So maybe, ultimately, these product related ads will will have to have a buy button on them.
Speaker 3:And so that you get your 4%, and then you can stack your CPC or CPM on top of that. Does that so any... Anything you wanna add to that? Or or, you know, do you think that's... Am I...
Speaker 3:You know, feel free to say, Scot, that's crazy. That's not how it's gonna go.
Speaker 1:No. I think I think that that's an obvious that's an obvious starting point.
Speaker 3:Today, if you're in Google AI mode, you get this terrible at the very bottom of the results, this little strip. And, you know, I think we can all agree that's gonna go away. That's just like a throw in. Well, we had it. It's like a Google Shopping ad thing that they've had forever.
Speaker 3:No one's ever going to see this, and this doesn't really generate any revenue. So that's kind of ridiculous. Now, this is when they announced in January 11 at, NRF when they kind of famously announced on stage, they did introduce an ad unit. And it's called Direct Offers. And they kind of conveniently show it's kind of a funny talk track because they're like, Let's say you're a retailer and you want to offer someone 20% off.
Speaker 3:And I was like, said no one, no retailer ever. So they love to have these examples where they're offering a very generous discount. But this example, if you recall, the consumer is looking for a rug, and then you can once it gets narrowed down into a single merchant, that merchant can inject an offer that is a deal, or they call it a direct offer. And then I took note that they were really big on talking about this almost kind of like, you know, I don't want to call it a store because it gets confusing, but it's a merchant area. It's kind of like the seller page on Amazon or everywhere where, you know, let's say let's say you're you're you're on AI mode, in Gemini, and you...
Speaker 3:They showed an example of Lowe's where you can then go into a Lowe's only experience. So you're looking for dishwashers, and you realize, well, I really wanna get it from Lowe's because there's one a mile from my house, I had a great Lowe's experience. You can then kind of flip it and then look at the... Only the Lowe's inventory for dishwashers, pick yours, and then inside of there is a Lowe's, branded chat experience, but also kind of like a, on Gemini Lowe's merchant experience. And then I thought, well, that's going to be obviously, Lowe's is paying for that.
Speaker 3:So that's kind of an ad in a way, but it's an upsell. But then I think there's going to be a lot of opportunities for ads in there. So now maybe that gives Lowe's a surface area to go sell Maytag and GE some ads in concert with Gemini, or is Gemini going to sell? So that's another kind of interesting surface area for us to think about.
Speaker 1:Well, I think that this that that's interesting in that it ties into to something I see a lot of potential in. This has existed for some time, but it's called collaborative bidding. And it's when a retailer and a brand both chip in to run an ad. You might have seen this on on social media or or or in Google search actually as well, where there will be an ad for that Maytag dishwasher. It will be from Lowe's.
Speaker 1:The ad will show up from the Lowe's handle, but Maytag shipped in for that ad buy and that ad unit, and it was based on, again, that retailers off-site audience extension data. So this is actually a pretty cool opportunity at a new surface for that collaborative bidding to continue by a retailer and a brand. That's a win win scenario because the retailer gets the sale, the brand gets the sale, their interests are really aligned in that.
Speaker 3:Is there ad tech technology for that? Like to do that collaborative bidding bidding? Is there some kind of like a manual thing on the back end that they they kinda sort out?
Speaker 1:Yeah, that is it. It's done. Some retailers manage it themselves with a team. There was a capability that DoorDash acquired a year ago called Symbiosis. I'm I'm I'm not sure exactly if that's still out in the wild.
Speaker 1:And then there's another there's another ad tech company called AMPT, which is getting started with this as well. Not in the Agentic space specifically, but
Speaker 3:I knew you'd know the answer to that. I know I did. I've already learned so much. This is amazing. Okay.
Speaker 3:So so Google has kind of planted a flag with one very clear one, and then I'm kind of anticipating there'll be some more on that surface over there on the kind of like the... What do you think we should call that? Like, should we call that the Agentic hosted merchant page, or should we call it their their hosted shop or something? I don't know. Do you have
Speaker 1:any At Agentic storefront?
Speaker 3:Agentic storefront. Yeah. Alright. Boom. We'll work.
Speaker 3:Heard it here first.
Speaker 1:We'll the we'll get the brand people on that one.
Speaker 3:This is how this is how, like, we we set trends here. And then this one was kind of buried. So, the lady that, that, you know, so, so I am 98%, I've met her, and I'm 90% sure when I met her, she was only doing Commerce, and now she has Ads and Commerce. So I think that's new. And she did a social media post that I thought was really, really interesting.
Speaker 3:And it's kind of exciting because I've been trying to get Google to do something more in Commerce for like literally twenty years. So first of all, she had this really cool little diagram that I don't really know if she she she hand drew that, but it kinda felt like it. And, you know, in there, she kinda shows that they have YouTube, Search, and Agentic Commerce. Those are her three priorities for the AI era, which is interesting. I was just giddy to see.
Speaker 3:Even though we're number three, I'm just happy to see Commerce kind of on on the napkin and Google. And then and then she talks about, you know, some things inside of there of how you you tie all this together, and, you know, there's ad surfaces across all this stuff. In that post, kind of buried in there, they kind of said, Well, here's like yet another ad format. And there's a long I'll link to it in the show notes, but she goes to a very long thing about they've actually been testing tons of units and they found that this one works really well. I haven't seen this in a while and I run these a lot, so I'm a little...
Speaker 3:I don't know what they mean, but this one reminds me of PLAs, so product listing ads in our old world where you could kind of like... You'd send a shopping feed and you could buy your way up in the results, basically. So, you know, the algorithm would kick in, and if you were the same price and things, you basically could could kind of buy your way up. And that's what it feels like to me. So here they show someone looking for, and there's a whole search prompt that you can go into in her post again.
Speaker 3:And, but here, here in the middle thing we're looking at, someone's looking for some headphones and they do a commute and they want to listen to it like on a subway, And it has to be noise canceling, and they they list all these things. So in there, we get down to a SKU, and this is the Sony WH CH seven two zero n wireless over the ear headphones, and it's offered by... I presume there's more retailers, but they just show Target and B and H, and then they showed that it's clearly sponsored. So I imagine there's gonna be a little sponsored section of retailers and then an unsponsored section of retailers on the offer, area of Agentic Commerce. So so this was interesting.
Speaker 3:So this one's... You know, they're they're kind of definitely crossing the chasm into, you know, you can buy your way up. And and they didn't say in this thing... And here, they actually have different prices. So so I'm not even really sure if you're just gonna have two sections, a paid section and an unpaid section, how this is gonna look.
Speaker 3:But but this is another... This is a second ad unit that we know Google's working on. There's the the deal one, and then this... I'm gonna call this an Agentic PLA for lack of a better word. How do you...
Speaker 3:How does that suit you?
Speaker 1:So this... I'm not sure if it... How how... This is actually the retailer bidding on being like a top retailer for a specific SKU.
Speaker 3:Correct.
Speaker 1:Rather than it being like an ad for a SKU.
Speaker 3:Correct.
Speaker 1:So it is the re... The... It's sort of like in in the Amazon poll on. So it would be a seller competing for the buy box and paying to get the buy box.
Speaker 3:Mhmm. Yes.
Speaker 1:Which we have never had before to my knowledge.
Speaker 3:Yeah. Because Amazon always wants the algorithm to pick who's in the buy box. And then once you own the buy box, you can run the ad. Right? So they've definitely crossed the chasm here in an interesting way that I thought no one, no one talked about this.
Speaker 3:I thought it was pretty wild.
Speaker 1:It is interesting. Okay. So this is, I'm getting slightly out of my depth here, but there is a really interesting, a critical part of bidding logic that different platforms bid it bring in, which is relevance. So if you you're selling these noise cancelling headphones, but someone searching for sports, you know, jogging headphones, not really the same thing. But if if someone searching for the jogging headphones and you're these noise cancelling over ear headphones, if you really, really, really wanna get in front of that buyer, you're gonna need to bid a lot because Google, for example, they have this relevance engine, which deprioritizes anything that they don't believe to be fully relevant as as an ad result.
Speaker 1:So you can buy your way to the top, but if you're not relevant, it's gonna cost a lot more than if you're actually a relevant product because they still want the ads to be relevant because ads that are irrelevant are really annoying. So wonder if even if you're, you know, in that example that you shared, B and H Photo and Target are at the top. They are very reputable retailers. If I'm some, you know, fly by night dollar store kinda retailer just popping up, I wonder how hard it would be for me to buy my way to the top of that retailer list.
Speaker 3:What what Tmall does... So so this reminded me of Tmall. So so you have... In China, have Tmall and Taobao. And a lot of your results are this kinda like, you know, enterprise kind of people and then, like, the the crazy, you know, entrepreneurial folks selling things.
Speaker 3:And they're always segmented very clearly into... And Tmall is always on top. So I kinda... It kinda felt like that. Like, maybe if you're one of the...
Speaker 3:If you're like a random eBay seller or an Etsy seller or something that happens to have these headphones, you're you're not gonna be in that top bucket. But then can B and H pay more and show up? Like, is it gonna be like a a a SKU level auction that's gonna go on? You know, what... I I can't wait to see this one in live and play with it and kinda see what's going on and how you...
Speaker 3:With PLAs, what you would do is you would send a data feed that had a bidding thing in it, if I recall. And like how much you're willing to pay per SKU kind of thing. So it's gonna be really interesting to see how they implement this one.
Speaker 1:Yeah. It it is... This is the constant balancing act that everyone's noted is this balancing act between trust and monetization. Because if, you know, some random retailer is showing up as the first option to buy these headphones, and I have actually seen this before with some, like, random eBay listing just show up for something I'm searching for in in chat GPT. And then that does make my experience not quite as good in the app even though it is a pay...
Speaker 1:That's a paid placement. So I think that that this is the crux of this whole question is how can the LLMs monetize all of this attention and all of this trust without killing the golden goose, which is the trust itself.
Speaker 3:Yep. This is the whole reason Anthropic ran those ads. I think it's going be hard and they know it.
Speaker 1:Let's come back. Let's come back to that because I've I've got some more thoughts on that one.
Speaker 3:Oh, okay.
Speaker 1:I didn't think they got it quite right.
Speaker 3:Spicy Kiri. All right. Yeah. And then this is one, I'm pretty sure you've seen this one. So, and so Rufus has an ad unit called a sponsored prompt.
Speaker 3:So so you... We've all seen, and Perplexity, I think, innovated this where you'll you'll you'll start off things with a prompt that used to be you had to start the conversation, but then they would have, like, little suggestions to keep the conversation rolling. And those were kinda like the suggestion prompts. And so then, a lot of them now, as they've gotten more and more users, they will give you a starting prompt as well, like some ideas of, like, I actually just had to re sign up for Claude for work, and it had three things like, do you want to think about a new product feature? Or, you know, they give you these little getting kind of icebreaker prompts, I guess I would call them.
Speaker 3:So there's like icebreaker prompts and follow-up prompts, which is a very common surface area in these chat experiences now. So what Rufus has done is here in the icebreaker area where it says customers frequently ask, they have put a sponsored thing. So it knows I'm looking for a TV, and now it's saying, which Samsung TVs are good for gaming? So Samsung has kind of paid their way into the prompt surface area with a clearly demarked kind of aspect. And then now if I hit that, now I'm just talking about Samsung TVs, which makes Samsung happy, and that's why they're willing to pay for that.
Speaker 3:Another one is and these are actual examples from their announcement of these things. Someone was looking for organic cereal, and in there, there was a, an ad. So one of their traditional type of... What do you call these? So this is a sponsored, carousel...
Speaker 3:Listing carousel, I think they call it. I don't know that they use the word carousel, but that's my language. Yeah. And in there, it's kinda like... So Nutrition and Co has done a brand buy for search on Amazon for organic cereal.
Speaker 3:So so we're in kind of the old world still. But in there, they buy a sponsored prompt that says, now I would to pay extra and have Sparky see if the customer wants to have you know, start a Spark sorry, Rufus start a conversation about this cereal. So I thought that was an interesting kind of bridge between the old ad model and the new one. So the old ad model is search banner takeover with products in it. And then a sponsored prompt kind of tied to that to get, if they're gonna...
Speaker 3:If they ever go into Rufus, I want them to kind of go into Rufus from my brand perspective, which is Nutrition Co. So then you click on that and now you're in this very sponsored kind of conversation about it.
Speaker 1:And that was the Perplexity that was the Perplexity model as well initially. Right? That was a sponsored prompt model.
Speaker 3:I have never seen one. I knew they were running an ad test, and, you know, to my knowledge, I haven't seen any live.
Speaker 1:Yeah. It got it got killed off because it didn't didn't perform.
Speaker 3:Do you know if Sparky... So several people have told me Sparky has them and I can't find anything. I can't find any examples.
Speaker 1:I... I'm not sure. I I don't... I I haven't heard anything about that either, actually.
Speaker 3:But I think this sponsored prompt thing is really interesting. So so here's an example. Ethically, would, so let's say I am looking at a Samsung TV. Would would would Amazon sell that surface area to LG to kinda come in and say, would you like... You know, do you wanna compare this to an LG TV?
Speaker 3:So so could could you do a competitive target with a sponsored prompt? I think that's interesting.
Speaker 1:I've also seen straight up sponsored products in Rufus results as well besides just sponsored prompts.
Speaker 3:Like they're mixing in the networks just kind of using the Rufus surface area in a traditional sense. Yeah. And then the other question is, if you're LG, do you ethically wanna do that? Like, you know, a lot of... In the olden days, we used to see this on Google, and then Google stopped it where you could go buy the the brand name, now now you...
Speaker 3:Only the brand can own the brand name within... With some some exceptions. Like, you know, three retailers could go buy... If you're a beauty retailer, you could go buy the... You know, they can all buy the e.
Speaker 3:F. Brand. But the... Do you think do you think that's something that that brands will cross that chasm? Brands don't care as long as they sell more, or do you think they'll get worried that they're specifically targeting a competitor like that?
Speaker 1:Well, right now, how... From a meat from a buying perspective, you can't just target sponsored product ads in Rufus. So I actually have I'm I'm I'm working on a piece right now. I'm doing a bit of a deep dive talking with some agencies about what they're seeing with regards to the performance of these roof responsive products. I'll have to come back to you on that one.
Speaker 3:All right. Future podcast. We'll have you back in ten years, which will be
Speaker 1:two weeks from now. Yeah. Next month.
Speaker 3:So that's kind of where we are. And then I do think it's not lost on me that the person running this at OpenAI is Fiji, and she built the whole Instacart program. So I do think we're going to see a lot more of this. So if you look at the Instacart ads, there's like that's just crazy. Like I could do a whole slide deck just on their ads.
Speaker 3:But they do lots of things in the discovery aspect. And then even at the product layer, they've gotten really good at tapping into and you know more about this. You and Jason know more about this world than I do. But there's these trade dollars, and this is the the big pool of money that that exists out there in CPG land that used to go in these circulars. And, you know, they've they've gotten really good at tapping into that.
Speaker 3:So here, they they they have, you know, at the bottom there with that yogurt, I think it's called a sponsored BOGO. So you can kinda, like, put an offer. This is kinda like that Google in in a weird way. Yeah. And then over on the right is a sponsored coupon.
Speaker 3:So you can actually kind of sponsor a product to have a coupon, and this one has a 75¢ off on some salsa. And then they have, you know, the one I've always experienced, I think, is a cool value add. As as I'm in the checkout experience, it kinda says, hey. You have one granola bar in there. If you if you wanna be in this Quaker Oats family, if you add two more things, we'll cover some of your fees for the shipping or something like that.
Speaker 3:And I think there's kinda like a cart kind of an ad format as well. Do you think we'll see... So my... Part of my theory is we're gonna see just this explosion of potential ad formats in the Agentic commerce world, and it's gonna look like some of these things will kinda come into play too.
Speaker 1:Yeah. Absolutely. And I love where you're going with this because I think a lot of my initial conversations with retailers, there was concern that it would immediately be a race to the bottom with Agentic Commerce and AI enabled commerce in that. Okay. In...
Speaker 1:If there is perfect competition. Right? If we know we can we can compare this offer from this retailer all in with shipping, with taxes, with the return policy versus that retailer that I'm always going to choose the retailer with the best price on that item. And that is not the, that's not the case. We shop at different retailers for different reasons.
Speaker 1:Some of that, you know, there's there's so many dimensions that retailers have always competed on, not just today. Selection, location, convenience. Do they... I spoke with someone recently who bought a microwave, lives in New York City. She bought it from a from a regional retailer called PC Richard because they came and picked up the old microwave and took it away and Best Buy doesn't do that.
Speaker 1:So that that there's so many ways that retailers can compete loyalty programs, sampling like with Sephora. People love the Sephora samples and it costs them absolutely nothing and it costs the brands, the brands fund that basically.
Speaker 3:Speed of delivery is another one like some of these, I'm not someone that would use the DoorDash delivery at Sephora and Ulta, but there are ladies in my family where they're having an emergency. And we're having a makeup emergency, and they will pay for that, or our dad gets to go deal Yeah. With
Speaker 1:there are
Speaker 3:a lot of more dimensions. And it's kind of like I kind of think of it as the service element, right? So the way you differentiate is you add layers of service. And it can be loyalty. It can be free add ons, and it can be getting the product.
Speaker 3:And a lot of another one that's really popular is payment mechanism. So I have some 20 year olds, and they love buy now, pay later. I would never use that. It feels kind of, you know, it's not my thing. But if you look at the data, there is a cohort of people that they don't like having a credit credit cards are for emergency, and they don't hardly put anything on there.
Speaker 3:And they like to know, I'm going to get I'm going to buy something big, and I have a plan to pay it down in three or four payments and be done with it, not have to stress about it. So so I do think there is gonna be a lot more levers that... I think it's actually gonna be a golden era for for trying to, you know, do marketing that we restricted to four keywords and and, you know, a CPC thing in the Google era. So I think it's going to look more like marketplace explosion of innovation around these things than what we've seen before.
Speaker 1:I hope so. I hope we see some new and creative ad formats. I also hope we see some new and creative targeting options as well because, you know, the the current the current MO is, I'm using all of your browsing history and information about what you clicked on, what you might be in market for, rather than the context of the chat itself and and the content that I'm looking at right now. And so this is where I wanna come back to the the Claude Super Bowl ad is that the reason why those ads were so good is that the the AI character was serving up a a... An endorsement of something that really didn't have any con...
Speaker 1:Context about what that person was really asking about.
Speaker 3:It picked up some context, but it missed all the other ones.
Speaker 1:Exactly. Exactly. So I think that that is a real misnomer. Of course, these like more intelligent models know what our our our rationale is. 99.9% of the time.
Speaker 1:Sometimes they get it wrong. But that that's that's the missing thing that I think we're gonna be able to get potentially with talk with targeting. I really hope that the the models figure out how to do this, which is target the context of the query, not who I am as a shopper or these like attributes or past past purchases that can be added on. But one of those caught ads where, you know, how can I get better at the gym? And it was something about him being short.
Speaker 1:So that's not We have
Speaker 3:short king stand tall.
Speaker 1:That's right. I love that.
Speaker 3:The writing on that was superb. Yeah.
Speaker 1:Yes. Fantastic. I I... The the really good ads, but I think what... Well, what I what I hope we will be able to get to is this more sort of content based contextual advertising rather than, some of the creepiness factor that comes from ads is like, oh, they're following me around.
Speaker 1:And when when they're too when they're too following us too closely, it feels invasive. So if there is a, you know, these and the only way that these LLMs will be able to do it is is purely algorithmic, which has opens up its own challenges for advertisers who want more control over their ad spend. But if we're able to sort of only bring an ad into an environment when I'm really like working through something or really trying to discover something, rather than like some kinda creepy thing reminding me that I have wide feet. Yeah. I don't need to be reminded of that all the time.
Speaker 1:Randomly simple.
Speaker 3:Asking for a friend.
Speaker 1:The I've talked about it publicly before. I've I've I've I've come to a... I've
Speaker 3:come to I've always wondered why Jason calls you duck foot Kiri. Now I now I know. Manny's mean. Dang dang it, Jason.
Speaker 1:He's so nasty.
Speaker 3:It is. Yeah. Yep. When you get to be the the head of publicists, all the things he does, that's that's how you roll.
Speaker 1:If only I had so many words in my job title. Just saying.
Speaker 3:I know. Yeah. But you have breakfast in yours, which is kinda cool.
Speaker 1:Yeah.
Speaker 3:Let's say opening up... Sam Sam Altman calls you and says, Kiri, I'm really unhappy with where our ads are going. I want you to come in and run our ads for us. Based on all this knowledge in your head, how would you build the ad program at Like what do you think would be the what are some ad formats you would put in there? How would you've already kind of hinted that you think this off-site opportunity.
Speaker 3:So that would be on your roadmap for sure, I'm sure.
Speaker 1:That's on my roadmap. Yeah.
Speaker 3:What else is on the curious queen of ChatGPT roadmap?
Speaker 1:Yeah. So definitely off-site audience extension. So you're not cluttering up the experience with with ads. Also, an opportunity to create new formats of ads like we talked about. Also, multi modal videos, try ons or I hesitate to say voice commerce because that didn't do so well last time around.
Speaker 1:I think I'm talking to chattypu tea more and more. But I think there is, right now, our experience with these tools is very much text and chat based. But as the the infrastructure gets built, I think there's such an opportunity for video as well. There is an amazing new feature in the Zara app, the clothing brand Zara. And if you upload a picture of your face and your body, it creates a a lifelike digital twin of you.
Speaker 1:And then you can do all of these virtual try ons and it looks just like you and you can see like this suits me or should I pair this blouse with this skirt or these shoes which looks better. It is a delightful experience. And I think that there is a real opportunity to bring in some really new things that we haven't seen before. Could we please throw away the sponsored product ad textbook and try and imagine things from a blank page because I didn't find that hot sauce sponsored product ad particularly enticing or alluring. I was like, come on, can we can we do better than this?
Speaker 1:So that that's my hope and dream is that we see new ad formats. We see sort of better ways of targeting people that don't feel creepy, that feel genuinely additive. And these ads can be helpful in the right place with the right targeting. I love at, you know, the ads on loves a strong word, but like Instagram ads. Instagram knows me pretty well.
Speaker 1:I've got a friend Brie. She loves TikTok and she's like, if I could give TikTok a monthly budget for makeup, I would because TikTok knows me so well now. And I actually trust that algorithm so well that I just kind of end up buying whatever it puts in front of me. So I don't know that there is a possibility that we could end up with something really unique and additive to our experience. I'm an eternal optimist.
Speaker 1:That's what I'm hoping for.
Speaker 3:Awesome. That's a perfect way to end it. Thanks, Kiri, for for hopping on in a kind of short notice and brainstorming this with me. You've given me a lot to think about, and I'm sure listeners and viewers as well.
Speaker 1:Thank you so much for having me.
Speaker 2:Follow, rate and share to stay up to date with Retailgentic, your guide to the future of AI shopping agents.