# S3 E7 T62 - Sharon Gee - Commerce

S3 E7 T62 - Sharon Gee - Commerce 

[00:00:00]  

[00:00:01]  

[00:00:08] Tom Limongello: Hey, Scott, you back from Ramp Up? 

[00:00:13] Scott Messer: It was ramp up was really great. I gotta tell you, such an amazing time with the greatest density of data professionals, uh, ever, and it's, it's a really fantastic 

[00:00:24] Tom Limongello: I, I definitely felt the fomo, not, not going, um, my calendar's pretty full right now. Going down to the Commerce Media Brand Summit in Atlanta on Monday, and then Marketecture with you. Um, then I've even got a new, uh, event that I'm going to the point of care, uh, POC now, uh, summit, which is in the pharma space. 

[00:00:45] Tom Limongello: So it is a busy spring. But yeah, data is really important and I think actually this episode that we're gonna be talking about, um, how agentic commerce is coming into the conversation, um, it was [00:01:00] something that I think we've held off in talking about. Um, but it really is a, it's a data discussion and it's sort of an infrastructural, uh, shift that is reminiscent of the e-commerce world. 

[00:01:13] Tom Limongello: Um, and the guest that we have, Sharon Gee, from uh, commerce, is going to talk about the fact that it's not as crazy as it sounds, the. Old infrastructure that you needed to be in Google Merchant Center is kind of the same thing that you need to be in these new agentic protocols. I think that's a good grounding for how things work, but she's a really good, she's good to spar with, and I think you were talking about who wins and who loses, and there's some really interesting parts of that towards the end of this episode. 

[00:01:42] Scott Messer: It certainly was, and I, I, I took a little bit of a pessimistic view into this and I wanted to talk about the winners and losers that we see as we shift. And she did a great job explaining sort of the. The beginning of commerce and why feeds were important and how they, how they really framed out [00:02:00] like the Web 2.0 that we've been in. 

[00:02:02] Scott Messer: And as we move into this new era, and I'm not gonna call it Web3 

[00:02:06] Tom Limongello: Please, please do not. Yeah. 

[00:02:08] Scott Messer: Uh, as we move into the agent era of browsing and commerce, I had a lot of questions about like, how does that rewrite the relationships that brands as advertisers and, uh. Websites, consumers, merchants of record. 

[00:02:27] Scott Messer: Like how. 

[00:02:28] Tom Limongello: And, and retailers. And for me, multi-brand retailers was another one where, what is, you know, and, and it hasn't been written yet, but it's, it's, it's, they're drafting it, 

[00:02:37] Scott Messer: I think if you ask Sharon, like it's been written, we're just about to like, we haven't just gotten to that chapter yet. Uh, so that was really fascinating. And I gotta say, one of the places where I really pushed hard was on in the loser section. And I was intimating that, um, retargeting of users was going to be greatly impacted because in ag agentic browsing, they're, [00:03:00] they're not coming to your site to, to do things, but you will get them as a consumer if they actually buy something, but you won't be able to retarget them. 

[00:03:08] Scott Messer: And so that may have a big impact on some of our favorite companies, particularly like a French well-known retargeting company. But what I. 

[00:03:17] Tom Limongello: think, I think that they would be very, very happy to hear that retargeting is, is dying because they've been trying to change that narrative, uh, when they moved into commerce media and now it sounds like they're moving into this agentic world 

[00:03:29] Scott Messer: And that's what I, that's the update that I have from when I talked to Sharon and reading the news. Now seeing that Criteo is powering some of open AI's ad tech that they're going to use for their, their ads beta that they're running. Seeing that, you know, it's probably not like what we used to think of Criteo, but it's most of that IPON web acquisition and building ad tech for companies and being able to create solutions. 

[00:03:53] Scott Messer: So without having done a lot of research and you know, we should get I, uh, Criteo and IPON Web back on here to [00:04:00] talk about it. I think that this may be their greatest opportunity and a great justification of like the skills and tools that they've been assembling and like they're meeting the 

[00:04:12] Tom Limongello: Well, they also, they, they also need to be given some credit that they went as deep as they did on Google Sandbox and didn't get much for that. So. 

[00:04:21] Scott Messer: Absolutely. Um, so it was a fascinating conversation. I think our listeners will really like it. Sharon is concise and precise, uh, and you're gonna love some of the metaphors. And if you like skid steers and tractors, like this is definitely your episode. 

[00:04:35] Tom Limongello: Yep. Yeah, I've already started editing this episode and, uh, when she gets animated, she gets full screen because she gets, she gets into it and she's got some great examples. So, um, it was great talking to her. Um, and yeah, I mean, this is, this is the beginning of a conversation that's gonna, that's gonna reappear on the Middlemen. 

[00:04:54] Tom Limongello: So, uh, let's get into it. 

[00:04:58] Tom Limongello: [00:05:00] Hello, welcome to the Middlemen podcast. We are here today with Sharon, gee, did I pronounce that correctly? 

[00:05:07] Sharon Gee: That's right. Yeah. 

[00:05:08] Tom Limongello: She is SVP of Product and AI for Feedonomics at a company called Commerce, which may sound like you might've heard of that, I know that Big Commerce is part of your, your whole organization, but tell us what this company is that you work for first, before we get into any of this. 

[00:05:23] Sharon Gee: Yeah, so little bit of background. Um, I joined BigCommerce six and a half years ago, uh, to lead the relationships with the channel partners, Google Meta TikTok, Amazon as part of that company, I hope. Lead the acquisition strategy for a company called Feedonomics, which is a market leading data feed management platform. 

[00:05:41] Sharon Gee: Because merchants have good data and they need to send it to all the sales channels, they have to drive discovery in this kind of omnichannel world at the time. And um, so what you, we acquired a company called Makeswift, which is a next JS React frontend visual editor. And so with these three brands, BigCommerce, [00:06:00] Feedonomics, and Makeswift we needed a parent brand commerce. 

[00:06:04] Sharon Gee: Is the through line. It's the parent brand that holds the brands, BigCommerce Feedonomics and Makeswift, which are all part of the company. 

[00:06:13] Tom Limongello: The reason I think we, we got connected was I was, I was at NRF and I was talking to one of your colleagues, Matt, and, I think that what was interesting for me was we kind of waited, we didn't really go headlong into LLM surfaces for advertising because it was very loud and everybody was making noise about it, but it was a lot of protocol talk, and it wasn't, it wasn't that interesting to be honest. 

[00:06:35] Tom Limongello: Um, and so it, when I had a, a great chat with him and he introduced me to you, I thought it was a really good time to sort of. Talk about what does it mean to be the merchant of record? What are the requirements for that? It seems to me, and you know, Todd and I, the other co-host on the Middlemen, we came from a background, uh, working in the Shopify ecosystem. 

[00:06:54] Tom Limongello: So we saw firsthand how smaller merchants. Who don't have a really good handle on their [00:07:00] first party data. Um, they need these services like Feedonomics to get their product catalogs in there. Um, you know, in this world, your company and others in this space are getting to be super interesting and important as a  

[00:07:13] Tom Limongello: lms start to have protocols that say, I need a merchant of record to pull products into a chat interface. 

[00:07:21] Tom Limongello: So tell us about what it's like working in that environment and what's happening. 

[00:07:27] Sharon Gee: Maybe a little bit of background so it's not new for third party channels to need catalog data. Google, in order to run product listing ads has needed catalog data for like a decade, right? Same thing with meta. And TikTok and Amazon, right? If you wanna list on Amazon, you have to provide them a bunch of information about the product, title, description, image, bunch of additional attributes to list on a marketplace. 

[00:07:53] Sharon Gee: So let's think about it. In eras we had like kind of the advertising era, whereas a merchant, if you had an e-commerce [00:08:00] platform, you could sync your catalog data to Google. And they would start to drive discovery of your products with a product listing ad or organic listings. Because when you're searching and looking for information, Hey, I want to, I'm looking for a linen dress in a size 10 from this brand, that information has to be provided to the search engine. 

[00:08:16] Sharon Gee: So you start with search, right? So feeds is this kind of anachronism of send my data to the channel so it knows about my products. That started with Google and with Meta and, and Microsoft ads, all of these advertising programs. That same data then powered the social shopping kind of era. And then the marketplace is era. 

[00:08:36] Sharon Gee: And now we're just into the next one where there are new channels who also need the catalog data. The Agentic era, right, where you have chat, GPT and Perplexity and copilot and Gemini. And they all need catalog data because they can't create a beautiful shopping experience where I ask for the thing that I'm looking for and they, the agent knows about it and can answer all of the questions about that [00:09:00] product. 

[00:09:00] Sharon Gee: Just by relying on scraped data, they need quality data because when you're scraping, it might not have real time inventory and pricing, and then they're surfacing, you know, 4 0 4 links of products that are out of inventory. That's not a good experience. So these answer engines need feeds just like every other channel before them has in order to be able to answer shopper queries. 

[00:09:20] Sharon Gee: And now that all these eyeballs are moving, you know, faster than ever to these answer engine channels that all of us are using. They realize that that query volume requires good catalog data from merchants of record. And so that's really the role we play. We have always sat between merchant systems, right? 

[00:09:37] Sharon Gee: Their pim, their ERP, their e e-commerce platform, and took taken that data and then massaged it into the format that the channel needs in order to surface that 

[00:09:48] Tom Limongello: Beautiful carousels. Yeah. 

[00:09:49] Sharon Gee: When someone is looking for it. And so it's not, the problem is not new, but the data and the channels is new. And so you said protocols and so everyone like went crazy. 

[00:09:59] Sharon Gee: Oh my [00:10:00] gosh. Why are there all these new protocol announcements? And the answer is because when you collapse the funnel. It used to be you would, you know, buy a product ad on Google and then, you know, you would capture some attention. It would plop you on a product detail page or a product listing page, and then you, as a merchant would pray for two to 5% conversion rate on your.com. 

[00:10:20] Sharon Gee: That's been collapsed now. Now people are asking these answer engines for their, you know, for their queries, and they're telling them their life story. And so the context has to be surfaced. Before they ever even get to your site, in some cases they might not even get to your site anymore. We're seeing organic traffic drop because answer engines are providing the answer you need when you're doing product research. 

[00:10:39] Sharon Gee: So now you discover and buy, in some cases that product all within that same age agent chat interface, and that means that the merchants have to expose that data, not just the catalog data that was powering search, but they need to power the data that actually allows transaction to happen on that channel. 

[00:10:57] Sharon Gee: Tax calculation, shipping [00:11:00] calculation, the ability to route and order, um, all of that payment like a, a tokenized payment. All of that is really important. And what it means is there are new protocols required for the data to go back and forth between merchant systems and these channels. And that's the same thing we've been doing forever, which is sitting between merchant systems and the channels that can drive them growth. 

[00:11:22] Sharon Gee: It's just new protocols that we need to send the data across. So if data's the oil. The, the protocol's, the pipeline, you, it's just how does it get to the place that it needs to go? And so that's really what we're seeing is the difference. There's a lot of noise in the space around these new protocols. Ag agentic commerce protocol with Stripe and open ai, universal commerce protocol with Google, um, and others. 

[00:11:42] Sharon Gee: You know, the, the new payments protocols with PayPal and Stripe, we support 'em all. Just like we supported the schemas of the ads and the social commerce platforms before that. So I don't think people should get crazy about what the protocols are. We need to understand that there's a new consumer behavior that is seismic in its impact, but it, [00:12:00] but understanding what's happening allows the merchants to make choices and understand that ultimately the data's the storefront, the customer's, the channel, and agents are now customers too. 

[00:12:11] Scott Messer: Okay. I really like all of that. Um, but as the tinfoil hat member of the podcast, I got a lot of questions. Um. And I'm gonna, I'm gonna back up a little bit in your story of like, let's get back to the fear part of agentic commerce is going to wipe out businesses, right? And so why should retailers participate in agentic commerce and hand over their feeds? 

[00:12:44] Scott Messer: Let me give this more preface there. Every time that a seller of some sort has sort of gotten into these platforms, it has become more expensive for them, more opaque for them. And just one more thing to go through. So why shouldn't [00:13:00] retailers resist this and what data do they get back because they don't own the user? 

[00:13:07] Scott Messer: And we're gonna talk about some companies that like may not have as big of a role to play in the future. Here, I'll answer that question later, but like. Why should retailers be in this channel and bring their feeds with them? 

[00:13:20] Sharon Gee: Yeah, why play the game? I'm gonna contend with you. That makes a good podcast. Um, this is not a marketplace play. Agentic channels. The merchant is merchant of record, so I did not say that it's going to ruin businesses. What I said was that it was going to change where the traffic originates for search and checkout. 

[00:13:38] Sharon Gee: So that mean, it doesn't mean that your business is over. You are the merchant of record. You're the person who has to fulfill the order. Uh, last I checked, OpenAI is not saying, and now we're Amazon, they're not. They actually are saying, Hey, we need to partner with merchants who are merchant of record, where they own the relationship on the return, the exchange, the loyalty. 

[00:13:58] Sharon Gee: So why would a merchant [00:14:00] want to have their products visible on the channels where all the eyeballs are going for the same reason they've had it on Google for a decade, and it's because they need to be able to have their products. Referenced when somebody asks a question and someone's doing discovery, they're more akin to a search channel than they are to a marketplace channel because the merchant owns the customer. 

[00:14:22] Sharon Gee: It just originates on that channel, and the initial checkout happens on the channel if in the cases of checkout, but in most cases it's actually an affiliate link because they're scrape right now. They're scraping a site and sending you traffic. So my answer would be because it's actually the acquisition channel where all the eyeballs are going. 

[00:14:40] Scott Messer: Okay, I, I'll take that. Um, in a, in a parallel universe, when TV shows were able to be recorded by TiVo, everybody forgot what Channel Dancing With the Stars is on. I bet you I could ask you any show that you watch right now, and you probably can't even remember whether it's on [00:15:00] Netflix or Amazon Prime. To most degrees, and they just show up in UIs and things and they're there for you. 

[00:15:07] Scott Messer: So there's a little bit of like a loss of branding here, and the customer may not even really, do they really know where they checked out from, or do they just know that the package is gonna be there in two days? 

[00:15:20] Sharon Gee: Well, they certainly know whether they're buying Nike shoes or 

[00:15:23] Scott Messer: They know whether they're buying Nike shoes or not, but do they know if it came from Nike, Amazon Foot Locker or Walmart? 

[00:15:30] Sharon Gee: I think what you're high, what you're focusing on is not, is the multi-brand retailer and is there challenge and risk to a multi-brand retailer? Absolutely true. Yes. There is challenge and risk there, right? How do you mitigate. Or what, how do you, how do you compete in this world where the value that you used to be providing, which is discoverability of multiple brands in a space where that discoverability is now happening within an answer engine? 

[00:15:54] Sharon Gee: Yeah, there's a, there's a big change in dynamics there. And so when we think about like. If not all of the, [00:16:00] all of the companies in retail are created equal. If you're a branded manufacturer, you're manufacturing products, of course you're gonna want that product. Data on the channels, right? Like we mentioned, Nike, Lululemon, Estee Lauder, Dell. 

[00:16:12] Sharon Gee: Those companies really want their products discoverable across those channels, right? Because they need, it went, you know, when somebody says, Hey, what's the best laptop for my 9-year-old Who's going to school? They wanna be able to show up when that answer 

[00:16:24] Scott Messer: Yeah, but like how does it impact Best Buy? 

[00:16:27] Sharon Gee: Yeah. Well, I think what's really interesting that we're finding is that, as you can imagine, we see the same pattern of adoption when a new channel emerges, always those channels say, Ooh, I need breadth of SKUs in order to meet my query volume in my searches. Right. That's happening. And so they go to partner with the channels who have the most volume of SKUs first. Right. We're seeing that in terms of, uh, you know, kind of the first gates of, of partnerships that we're seeing publicly announced between folks like OpenAI and Walmart or others. 

[00:16:55] Sharon Gee: Right. So when you're a multi-brand retailer, your benefit, you're like, [00:17:00] Myers-Briggs personality as a brand is different than if you're a branding manufacturer. Your, your benefit, your strengths are about your breadth of capability. And so I think there's a huge opportunity that we have, um, you know, in that category of, of retailers to understand the data sets that you have and to be really powerful and, and deliver that data at scale. 

[00:17:19] Sharon Gee: Because in many cases, one of the biggest challenges that exists is. You know, there's inconsistent data that doesn't perform optimally across channels. But if you wanna be where the checkout happens, you are a merchant of record, meaning the agent places an order on your site. If you are the brand that has the best data on a specific category of products, and therefore are answering the questions best when they show up on open ai, imagine how effectively you can compete when you have comprehensive data across a category unlike  

[00:17:50] Tom Limongello: Well,  

[00:17:51] Scott Messer: Right, so, so they also have, so then one to the first part things like. Return policy shipping speeds. The soft, the [00:18:00] soft side of these retailers I think becomes very important. And you're also gonna have to communicate to the agents, and that's gonna be a big thing, but the customer doesn't really care where it comes from. 

[00:18:11] Scott Messer: But they may say like, oh, and make sure I can return it. 

[00:18:14] Sharon Gee: Or make sure I make sure I get my points 

[00:18:16] Tom Limongello: Yeah, my  

[00:18:17] Scott Messer: I get.  

[00:18:17] Tom Limongello: point there, 

[00:18:20] Tom Limongello: there.  

[00:18:22] Scott Messer: To the, to the other part, it's not just about who has the best, most complete data. Now you're gonna start getting into price and discounts and maybe loyalty points, right? So how does that show up in feeds? And I, I think this will lead into a question of like, how do we run promotions through feeds and how do we do all this merchandising 

[00:18:44] Sharon Gee: Yeah. 

[00:18:45] Scott Messer: through this because. 

[00:18:47] Scott Messer: Retailers know I can merchandise on my sites, I can write merchandise in a few places. It's probably hard to merchandise in like the Google product grid, but like how is it, how is this gonna happen in the l LM world 

[00:18:59] Tom Limongello: Or, [00:19:00] yeah. Or, or is it a better way of saying is is like. Nobody ever really cared about Google shopping. Right. And how does this turn into, you know, if this is really gonna succeed, how does it just not be another version of Google Shopping 

[00:19:14] Sharon Gee: Yeah. interesting when you talk about Google shopping is like as a, you're saying as an individual product, but Google search, a lot of people care about, 'cause it drives a really significant amount of volume. Right. Totally different. But the data that underpinned both of those experiences is actually the same. 

[00:19:28] Sharon Gee: It's the data that lived in Google Merchant Center. So when we talk about, I love this question, what data do you need to send to the channels in order to have them merchandise you in the way that you wanna show up? This is where we start. Um, we've thought about this problem really deeply, as you can imagine, because. 

[00:19:45] Sharon Gee: When we, we work with 30% of the internet retailer 1000, and they have the same question, which is, how do I get the answer engine to say my name instead of my competitor's name? Well, it has to do with the data because they're not, it's not a, a storefront experience you [00:20:00] own, right? And so what the information you are either making available for the answer engine to scrape from your own.com. 

[00:20:08] Sharon Gee: AEO / GEO, right? Is that practice or that you're externalizing in a feed so that they can ingest directly to their veins. So it leaves in the mind of the answer engine, um, and it they say your name. This is where there's this huge focus on data feed management and orchestration in this new era because. 

[00:20:25] Sharon Gee: Imagine you have an amazing, you know, pair of shoes that is really good for running, but it's not really good for travel or leisure or, uh, you know, some of these other use cases. Or imagine you have a laptop that has very specific specifications, including a few areas that would really indicate whether it should be merchandised in a set of queries around is this good for your 9-year-old who's looking for, you know, and doesn't have the parental features that you need, so all of a sudden your product. 

[00:20:52] Sharon Gee: Um, data enrichment strategy becomes really critical. You need to be able, you need to be able to auto generate all of the attributes that these LLMs [00:21:00] use to answer questions. You wanna be able to create question and answer pairs around, is this linen dress really good to wear in Italy when it's hot? 

[00:21:08] Sharon Gee: Right? Like, you need to be able to make sure that the context exists in the product descriptions and in the data that you're sending to these channels so it can self 

[00:21:17] Sharon Gee: based on 

[00:21:17] Tom Limongello: I guess that that story makes sense to me. However, it is a muscle that the brands haven't had to really exercise to that level. And to be fair, you know, like if you go into the retail media discussions that happen in a JBP situation, um, what the mismatch between the brand and the retailer is always on. 

[00:21:36] Tom Limongello: The fact that the retailer cares more about the category and the brand care cares more about their own brand and their own sales, and. What I'm interested in figuring out when we were talking about this in our prep call, um, was this concept of agent to agent. And in a world where there's agents connecting with each other, the one, you know, the, the entity with better data has a better [00:22:00] chance to, you know, provide better context. 

[00:22:02] Tom Limongello: And the retailer, knowing the whole basket or understanding cross category data and things like that, um, is going to perform better. Then, you know, a toothpaste brand that potentially doesn't have a lot of connection to the rest of the travel industry. Right? And so it's like, yeah, I need toothpaste for when I'm traveling. 

[00:22:22] Tom Limongello: Um, but it doesn't know about matador's, you know, toothpaste, containers, or you know, like, you know what I mean? Like, it's just not, it's not gonna be able to data enrich to that extent, whereas a rec retailer could, and maybe they need to sort of get better. Even at being retailers and merchandisers, but I feel like there's something missing here where it's like, yeah, you could force the brands to go crazy with this stuff, but they're still kind of like, you know, buy my product, not buy a whole set of products or a whole category of products. 

[00:22:51] Sharon Gee: Yeah, I, I agree with all of your assessments around the strengths and pros and cons of being each one, right? Because there's always been kind of this interesting relationship between the brands who make the [00:23:00] products and the retailers who sell them, right? Like that's, that's like not an. Thing. It's just we're in a new channel where that's the case. 

[00:23:06] Sharon Gee: And I think what's really interesting is I kind of use this joke about Myers-Briggs personalities of brands, because they all have their different strengths, right? You get to be, you can be fast, or if you're a manufacturer, you know the depth of that product, you know how it was manufactured, you know everything about it, but you don't have that category level information. 

[00:23:22] Sharon Gee: Whereas if you're a multi-brand retailer, you know. How this relates to the context of, you know, all the rest of everything that's being merchandised for Valentine's Day or for, you know, whatever seasonal, upcoming holiday is coming up. So it's like there's pros and cons of the data that you can put to work based on what, you know, like if you look at like Dick's Sporting Goods or REI or other, you know, multi-brand retailers, they become, they have become amazing experts at very specific niche areas where they can give you context around this one versus that one. 

[00:23:52] Sharon Gee: So if I was on the data team or the marketing team who is powering Ag agentic on some of those teams, I would be focused very much on what we've heard a lot of our customers say, [00:24:00] which is, I need to get my data in order. Because we know in this new world, the data's the storefront that how you externalize your data, whether it's on an owned property. 

[00:24:08] Sharon Gee: You know, that is being, you know, traversed by humans or now increasingly by agents. Um, whether it's an own.com or whether that's data you're feeding to your own agent that you're offering on a, on a.com experience, or whether it's the data that you're sending to, you know, these third parties that are driving them. 

[00:24:27] Sharon Gee: The, you know, discovery across, across the, you know, the whole cat catalog of retailers now, I think what we know is that the, the substrate, the most important layer is do, are you getting the data right? To be able to answer questions that a agents now are surfacing wherever the data's coming from. So, you know, that that's one of the areas that I think is the biggest focus. 

[00:24:46] Sharon Gee: And certainly you mentioned at the top of the call, I think why we're seeing so much interest in this category of data management and orchestration. 

[00:24:54] Tom Limongello: Um, well, one of the things, and the reason, you know, when, when Matt Crawford and I were talking, I was, I was [00:25:00] kind of, you know, wondering. For many cycles, like when I was at Quotient, I would sit in meetings at Albertsons where we were the, you know, the, the retail media partner and Google was trying to come in and take our business, and they couldn't, because Google wasn't in this transaction world. 

[00:25:17] Tom Limongello: They, they had tons of data, but they didn't have that final transaction or a loyalty program, to your point. Um, now they, and Meta and others. Could potentially be in this world. And you know, what has changed in this world, you know, with the merchant record, I think there's a lot of questions about, you know, Walmart's partnership with OpenAI and other, you know, sort of setups like are they giving away the store and allowing Google and Meta to have a bigger role in commerce now or is it still just another channel? 

[00:25:53] Tom Limongello: And you know, they're still sort of locked out of seeing the transactions. 

[00:25:56] Sharon Gee: I think it's pretty early to say. So here's, here, let me, let me give [00:26:00] you some analysis around that. Um, what we know that they're the ones with the ch with the eyeballs, right? If people are using the consumer apps, what are you gonna do? Not be on the channels where all everybody's actually discovering products you  

[00:26:10] Tom Limongello: Yeah, I think we all agree that this is, like, I use it all day long. Um, you know, I use it for harder products, not really as much for, you know, products I buy all the time, but I, I think there's real disruption's gonna happen, so totally agree with you there. 

[00:26:24] Sharon Gee: So let's talk about Google as an example. So Google has one of the best understandings of the, of the products of the world ever because they've been running product listing ads with Google Merchant Center data for like a decade, right? So they have it. 

[00:26:36] Tom Limongello: They know the fundamentals. 

[00:26:37] Sharon Gee: have a foundational understanding of all the products on the internet, and because they have the ability to send local inventory and realtime inventory availability, they actually know not only what products you have, they know where they're located. 

[00:26:49] Sharon Gee: And so that data set is really powerful. So when you wire that up with something, they also own Google Pay, which is a wallet that knows the consumer's identity.[00:27:00]  

[00:27:00] Tom Limongello: Yeah. 

[00:27:00] Sharon Gee: Because think, think about, and it's already pre-integrated to all the payments companies in the world on dot coms, right? What, what's really interesting here is we, we, we are bifurcating all of the data sets, right? 

[00:27:10] Sharon Gee: So you, it used to be that tax calculation, shipping calculation, order routing payment all happened within an e-commerce checkout, right? Like on an e-commerce site. Well, now. Catalog can be sent directly to the channel, which is, it has been the case for ads. But you can have that collapsed funnel where that checkout experience happens. 

[00:27:30] Sharon Gee: But they have to know the consumer. 'cause otherwise there's a fraud problem. So the companies with wallets who know consumers and merchants are in a really interesting position here, which is why you've seen people like PayPal and Stripe and Google start to issue new protocols. To the world that says, here's how we're gonna hand the data back and forth, because now this checkout is not happening on owned.com experiences. 

[00:27:52] Sharon Gee: It's actually happening ally, the cap. It's being captured within this age agentic experience, but the merchant remains [00:28:00] merchant of record. So we're gonna go process itally on their existing systems in this, in, in some cases to make sure that the merchant can actually fulfill or, you know, pro provide a return or et cetera. 

[00:28:10] Sharon Gee: And so that dynamic, that change, it's not an ads channel, it's not a marketplace. It's a third thing. And so when we talk about Google needing to release Universal Commerce protocol, we were one of the first partners that were announced the day they announced it. We, we saw the protocol and, and gave feedback on it early. 

[00:28:27] Sharon Gee: Um, you know, 'cause when you have tens of thousands of merchants on an e-commerce platform, in addition to sending feeds, we've got a pretty good understanding around what data is needed in order to be able to support that experience. And so universal commerce protocol is Google's approach to how they're doing AG agentic because they have really unique assets, they have catalog data for that's been powering ads for a long time. 

[00:28:47] Sharon Gee: Google Pay and they have a concept of loyalty and of, you know, they have business agents who already know about a merchant's brand. They can already talk about a brand and its capabilities on their own Google owned [00:29:00] experiences, right across YouTube, Google, you know, AI summaries on Gemini, et cetera. And so it's a really interesting ecos landscape change because if we compare them to say someone like. A new entrant in the market who doesn't have a Google Merchant Center to speak of, right? They don't have a concept of catalog data across the world, the way Amazon or Google or Microsoft. Do they need it? 

[00:29:22] Scott Messer: a merchant center. 

[00:29:24] Sharon Gee: I know. 

[00:29:25] Scott Messer: A merchant center. 

[00:29:27] Sharon Gee: They don't have a merchant center like they can consume that data and de-dupe what they're scraping and align it to, oh, this is the merchant ID and the SKU id and here's the transaction ID that allows us to, you know, pop up this, this button for checkout to happen. 

[00:29:40] Sharon Gee: So it's a really interesting dynamic because new channels who have the eyeballs. Aren't necessarily the ones who can, who have the data at scale. You talk about like Microsoft, they have Microsoft Merchant Center. People have been running Microsoft Bing ads for a really long time. They actually have a really comprehensive data set around around catalog that could make them a really interesting player, [00:30:00] which is why you're seeing them in this space. 

[00:30:01] Sharon Gee: So I think, and then when you think about partners like Perplexity, you know, we're powering ex, we're powering. Live Agentic checkout experiences on perplexity in partnership with PayPal and Commerce because they've gone and done a partnership really quickly that said, Hey, PayPal, you can get me what I need from a catalog and from a payment and a wallet and an identity and a fraud perspective, and then route the order to a merchant system and partnership with commerce. 

[00:30:24] Sharon Gee: Let's do it. So we're starting to see some really interesting partnership dynamics emerge because anytime you run a new protocol, the first partners who get to the protocol have the market. Then it expands because the protocols are public, right? So that, we saw that with the Universal Commerce Protocol. 

[00:30:39] Sharon Gee: We saw it with Agent Commerce protocol with acp. And so I think we're in this kind of funny early innings moment in time where eventually, just like Google Schema, you know, schema.org and, and the Google Merchant Center schema kind of became a standard that everybody adopted. We're still in earning early innings, but the channel partners and their ecosystem of partners are still kind of building the [00:31:00] rails on which the new data will need to flow to power these experiences. 

[00:31:03] Tom Limongello: So you've seen this happen a couple of times. You've been there six years. You probably saw, you know, a a, a fair number of these evolutions turn over from the perspective that you have here. What do you think happens next? Like, what are the, what's, what's coming? 

[00:31:19] Sharon Gee: Well, I think what's always interesting is when you start to consider complexity and fulfillment, right? So one of the biggest challenges that you have as a, as a. Um, as a merchant is the categories that are easy to buy with a thumbprint. I call them TikTok categories. It's things like consumer electronics, appar like apparel, footwear, something that fits in a box that's like this that can be easily shipped to you. 

[00:31:40] Sharon Gee: Let's imagine that you're a furniture. Retailer or um, or a manufacturer, it's kind of hard to make sure that those channels have all of the things that you need for your custom upholstery, for your white glove delivery, for, you know, the insurance that you need to have on the truck when it's delivered to that person's living room. 

[00:31:56] Sharon Gee: Like that  

[00:31:57] Tom Limongello: but this is, this is really relevant because these [00:32:00] are the types of things that I envision myself needing chat GPT to help me figure out whether I buy. 

[00:32:06] Sharon Gee: Yeah. And so that's why the, that's why it's a third thing. It's not a marketplace. It is both a discovery, it's it's both. It's both a 

[00:32:13] Tom Limongello: Well, yeah, but like when you think about it, I think we're all thinking about this in the old channels mo mode because Scott went straight to price comparison, right? And like, you know, my experience right now is like trying to buy Skai equipment or something. And I'm like, yes, price is a factor, but that's not the only thing I care about here. 

[00:32:30] Tom Limongello: If, if that were the case. I would just, you know, I would just buy the lowest price thing, but that's, that's not gonna satisfy the 

[00:32:37] Scott Messer: I'm not, I'm out here catching, catching strays from Tom over here. So it it, it's a good point though. Like, what is this? If, if it's not product feeds, how do you get like checkout structure into it to know like all the things that I need to really fulfill your order. Is that a, where does that come from? 

[00:32:59] Sharon Gee: I love this [00:33:00] question. If it's not the catalog data, how do you make sure you're surfacing the tax calculation for that 

[00:33:05] Scott Messer: The tax calculation, the color of the upholstery be like it's in there, but like we need a custom builder thing or you know. 

[00:33:14] Sharon Gee: So that's why when everybody says, oh, websites are dead. I roll my eyes because they're not dead. They're just changing. Guys, like you'll always have a website is the way a brand communicates who they are to the world. It's what they own. It's their ability to say, I have better tools. Imagine if you wanted to be able to go to a a Furniture Brand's website and talk to their agent that could custom design you that room for you based on all that information. 

[00:33:36] Sharon Gee: So it's actually an amazing opportunity. For these experiences, these branded experiences, to really morph, oh my gosh. The landing pages, they got so boring and every product landing page looked exactly the same. It was like, we have sameness forever. I'm so glad that we're to a place where the sameness has been abstracted now with an agent and the, and the uniqueness around what brands can offer will be visualized on these dot coms, [00:34:00] because that's  

[00:34:00] Tom Limongello: But their, their ag agentic capabilities are gonna suck for the short term. So like, is there a whisper protocol? It would be like, take all my stuff from chat GPT over, 

[00:34:10] Scott Messer: Or like a handoff, be like, you know what? I actually can't buy this couch for you. You should go to the website. 

[00:34:16] Tom Limongello: yeah, I mean like, 

[00:34:17] Sharon Gee: To be clear, like on these answer engines, if they're, if the data that you need to buy this product doesn't exist, merchants are able to elect in the data. Like I'll just, I'll just be really tactical. When you send a feed to these channels, you have the ability to say, make this one available for checkout. 

[00:34:32] Sharon Gee: Don't make this one available for checkout. I'm a furniture. I'm saying couch not available for agentic checkout pillow. Yes. Available for age agentic checkout. Right? So it's, that's all managed in the feed that's getting sent to the channel that says, enable this or don't enable this. Now for some channel, it's like, it's a little bit more butcher knife on the early channels where maybe they don't have those controls, but they're working very fast to fix that. 

[00:34:53] Sharon Gee: So I, that's like a matter of weeks. It's not even worth 

[00:34:56] Scott Messer: Have the, have the, have the companies ingesting these product feeds for ag [00:35:00] agentic things handed out sort of a spec or requirement for like, here's where we want. The question and answer things of, you know, is this linen dress nice in Italy? Like, have they said this is where you put it and this is what it should say. 

[00:35:14] Sharon Gee: Yeah, so it's depends on the channel, right? So as you can imagine, Google with Universal Commerce Protocol has done that. They've said, here's the way we want you to send catalog. Here's the merchant center fields that you need to be able to support. Um, so we, of course, the, uh, working very closely with them knew what those protocols were and implemented them literally overnight so that when when somebody says, Hey, I've been accepted to the Google Commerce, you know, to the Universal Commerce Protocol program, we say, great, your feed's already ready. 

[00:35:39] Sharon Gee: Go ahead and send it over 

[00:35:40] Scott Messer: But they need to go fill. They need to go fill in. All the information 

[00:35:44] Sharon Gee: well, they can or they can use our service that automatically fills it in based on all the data we already know about them. Right. So if you have all 

[00:35:50] Scott Messer: the dress is like a q and a of, is this dress good for Italy in the summer? 

[00:35:55] Sharon Gee: Exactly right, we can literally generate using. Like us and [00:36:00] others can automatically generate data based on all the things that we need or know about your product and what the new channels need. So we can automatically use AI vision to analyze the products and understand what the attributes are. It's a short sleeve shirt with black polka dots, and we can automatically generate that data into the, into the catalog and, and send it to, to  

[00:36:19] Tom Limongello: So that actually to me, 'cause I, I remember hearing the perplexity guy talking about, hey, ads are gonna be for agents and I don't think that's the case, but, but attributes are, and using computer vision to, to analyze images that are on a site and things like that makes sense to me. I just don't, I still don't really know The handoff, I think is, is the issue for me. 

[00:36:38] Tom Limongello: Like, who's gonna stop, you know, stop doing the work in their platform to actually go to the other one. 'cause right now, if you use one of these, a agentic browsers. It's a nightmare. Like, you know, it, it, it's like, I dunno if you've, you know, if you played with like open call or whatever, but it's like, I got through the first hoop. 

[00:36:53] Tom Limongello: I'm, I'm now going into the se, I'm going looking at a picture. I'm now, I looked at a whole bunch of pictures. I'm gonna go and I forgot to do this. Like, [00:37:00] it's a terrible experience. Um, and so I'm, I'm just trying to figure out like, how does this become smooth? And right now I don't see any, you know. 

[00:37:10] Tom Limongello: Understanding of how this becomes smooth. I think from your perspective, you're showing how things can be automated, like data can be filled and things like that. But I, I still don't really sort of like the rules of working. I guess that's what's gonna happen is here the protocol can't do all of the work. 

[00:37:24] Tom Limongello: I'm envisioning like a A to a JBP program, but I don't know if that'll ever happen, 

[00:37:29] Sharon Gee: Well, Google's already produced an A to a protocol, right? That eventually, like, so let's talk about where we are versus where we're going, Sharon, like it used to be a shopper to merchant relationship. Sharon would go to a website, right? I go to anthropologie.com looking for that dress. Now, I asked chatGPT, or I asked Gemini, Hey, I'm looking for a dress to wear to my friends wedding in Italy, and I'm looking for a size 10 and it needs to be a man of linen 'cause it's gonna be super hot in July. 

[00:37:51] Sharon Gee: And it says, Hey. And I say, and I usually like Anthropologie. Can you recommend some other brands for me that I should consider? And it says, Hey, here's some options for you. That's what we would call an an [00:38:00] agent to shopper experience. Very soon there will be a agent to agent experience where a shopper agent and a merchant agent talk and negotiate on our behalf. 

[00:38:08] Sharon Gee: But that doesn't exist right now 'cause merchants don't have agents yet, but that's what they're building. If they can, and they will. And a lot of, a lot of the, um, you know, very technology forward ones are starting to leverage the same technology to be able to build agents that know everything about them because they have the data. 

[00:38:25] Sharon Gee: They're the ones with all those product specialists who have YouTube videos who've said, why this product versus that one is so good and this one's good for that. And that. That's data the brand owns, and so they can put that together and, and train their own agent. So let's imagine you're a Caterpillar and you are a B2B brand and you sell, you know, heavy, you know, farm equipment and you know exactly what the replacement part is for that, you know, skid steer that somebody bought. 

[00:38:49] Sharon Gee: I am so excited about what, what brand agents are going to be able to do for the B2B ecosystem. If you're a B2B buyer and you say, Hey, I bought this tractor and I need a replacement [00:39:00] part, can you please find the one for me? Or, Hey, I need to be able to, you know, run a PO for these 120 items that I previously bought. 

[00:39:06] Sharon Gee: Just order them all. Again, those processes are so manual and terrible, and so it's why we've released a purchase order agent for B2B brands to be able to address that. So I think when we talk about agentic. There's like a agentic third party channels that all of us know every day and love, like chatGPT and Gemini and perplexity. 

[00:39:23] Sharon Gee: And then there's what brands are building that will be, that will be owned by the brand, fed by the data that only they have. And so when we talk about why would they participate, the answer is some of them might not. Because if you're a B2B brand, um, merchant, if you can't surface your unique customer price list within that shopping channel, you're not gonna wanna do it. 

[00:39:43] Sharon Gee: You'll wanna make sure that the catalog. Is there, so that when somebody's looking for that, ooh, they know that website over there has this product available. They're the ones who have replacement parts for the Caterpillar Skai Skid steer. But you might not surface pricing, right? You might surface it in a, or you might surface a list [00:40:00] price and know that that's where the real pricing happens, because it's about, you know, loyalty or customer entitlements based on, you know, your, your customer group. 

[00:40:09] Sharon Gee: So I'm actually really excited because ag agentic. Agent Agentic is an equal opportunity, whether it's on third party channels or whether it's on owned channels like the brand agents you're building. So that's why we're seeing a lot of the brands today say, I have this data. It needs to feed both my acquisition channels third party, but also the new agents that I'm about to build on my own experience. 

[00:40:29] Sharon Gee: And I think that that's where a huge amount of opportunity for your multi-brand retailers lives. Because imagine if I go on to REI's, uh, normal.com and I'm a loyalist and I have so many points that I wanna spend there, and their brand agent shows up and I say, Hey, I'm taking my kids camping for the first time. 

[00:40:45] Sharon Gee: My daughter just turned six. What gear do I need? Think how powerful that agent is in helping me buy a bunch of stuff. So anyway, that, that's why I think about that. 

[00:40:53] Scott Messer: I, I really enjoy this. Let me, um, let me bring back some doom and gloom for us here, uh, as the, as the. 

[00:40:59] Tom Limongello: [00:41:00] Uh, before you do the doom and gloom, I just wanna say thank you for coming on and, and having that you have the product catalog in your head somehow, so you have an ability to talk through these examples. You went from John Deere skid steers to Italian, Italian linen dresses to. Camping gear, so nice work. 

[00:41:16] Scott Messer: Uh, yeah, it very true. Uh, I would also like to take a picture of the skid steer and submit that and be like, that's the thing that's broken. And circle it with a red thing and be like, you, you a, I deal with that. 

[00:41:27] Sharon Gee: This is possible. Now the we, the future is here. It's awesome. 

[00:41:31] Scott Messer: here, but let's talk about who the future does not include, and who is this like bad news for four. 

[00:41:38] Sharon Gee: Okay. 

[00:41:38] Tom Limongello: I think this is gonna turn into a segment, Scott. bad news, like yeah. 

[00:41:44] Scott Messer: Bad 

[00:41:44] Sharon Gee: Well. 

[00:41:45] Scott Messer: I mean, there's, there's some companies that I, I agree with you that internet's not going away. Websites will still be around, but, uh, one of two effects will happen. Either a large percentage of your web traffic will go away and [00:42:00] become a agentic. And so you have to rethink your investments in traditional MarTech. 

[00:42:05] Sharon Gee: Yep. 

[00:42:06] Scott Messer: Um, that's really, I think, the thing that will happen. Uh. So in that, like let's talk about some areas that might have some exposure and risk to it, which are the middle people. Um, because people aren't visiting your website, abandoned cart and retargeting and even newsletter signup probably gets really difficult. 

[00:42:33] Scott Messer: Would you agree? 

[00:42:34] Sharon Gee: I think it depends. 

[00:42:35] Scott Messer: Uh, in an agen, in a agentic universe, they can still go to their site and do that stuff. But like in this agentic shopping universe, is there as much room for that? Do you think the agent will be like, would you like to sign up for their newsletter 

[00:42:50] Sharon Gee: uh, I do think the agent will say that. 

[00:42:52] Scott Messer: and not a paid ad? 

[00:42:53] Sharon Gee: Yeah, I think, I think when we think about what functionality do dot coms need, it's the same [00:43:00] functionality that's already existed. But where there becomes pressure, I, I think is, you know, people can vibe code websites like overnight. It's amazing. Like the, like we live in an, we live in the future. 

[00:43:10] Sharon Gee: It's so exciting. I used to, you know. Work at web agencies for like back in the day and it's like what used to take three months and $300,000 now you can do overnight and like Claude Code and one shot it, it's like amazing. But what I, what I think is like really interesting is when we think about those third parties that are currently like niche SaaS providers. 

[00:43:31] Sharon Gee: Where they're kind of like holding accounts and, and putting, you know, taking data from one place and putting it to another. Those are the ones that I think are at serious risk because the ability to get data from place to place has never been easier. It's never been easier to create, like, you know, it's a, it's a super communicative, um, ecosystem where agents can now build connectors between, you know. 

[00:43:53] Sharon Gee: App ecosystem is very, very straightforward. So I think the platforms are going to remain because they have a lot of data, right? And a lot of access [00:44:00] to, um, ability to surface that data or look at it holistically in order to identify insights that allow you to take action. I think if you are a niche software player who only does kind of one very narrow thing, the risk of you becoming a feature of a platform instead of a company is very high. 

[00:44:16] Sharon Gee: So that, that's, that's talking about it generally. You can talk about categories if you like, but I think that's how I would, I would characterize that 

[00:44:22] Scott Messer: Yep. We're gonna talk about some specifics, uh, because the, the ability for like retargeting, right? You don't really get to retarget. These users and there's a good example of like, well, if they're on Google shopping and they click and they see your, your, your product, but then they go to the other one, you don't get that user anyways. 

[00:44:41] Scott Messer: But that's kind of the point. So the people who are providing a retargeting solution for the brand or the retailer may not get as much opportunity to retarget, which would inherently be bad for the company providing the retargeting service. 

[00:44:58] Sharon Gee: Yeah, if they don't, you know, [00:45:00] change the way all the rest of us do in terms of what value that those companies offer. So I think if you're a software company and you see the new ecosystem, and let's imagine all of the new. You know, and, and to kind of counter your doom and gloom, there's brand new technology companies out there that never like, look at the, the category of like, 

[00:45:15] Scott Messer: I'm very hopeful for many things. 

[00:45:17] Sharon Gee: yeah. 

[00:45:18] Scott Messer: I'm very hopeful for many things. 

[00:45:20] Sharon Gee: If you're in the, if you're in the game of, I want to understand consumer traffic on sales channels, that's what I would say retargeting is you're in the game of understanding how are consumers interacting across sales channels. We've seen an entire new like explosion of these, of these LLM visibility channels. 

[00:45:37] Sharon Gee: Think about profound and scrunch and every other competitor that's like popped up overnight where they're simulating. Session data using persona specific queries and then, you know, doing a whole bunch of simulated sessions and licensing a bunch of data from all over the place in order to be able to say, Hey, we know those, these LLMs haven't delivered analytics APIs for us to give you yet, but here's what might be happening and [00:46:00] don't you need to optimize your data in order to be able to perform there. 

[00:46:02] Sharon Gee: So my response to you're like, yeah, if you say you're a retargeting company, yes, maybe there's a challenge for you, but if what you are actually are is understanding how to help. Merchants or customers continue to know who are their shoppers and where are they and get at them. You know, they like all the rest of us are gonna have to figure out how to operate in this world. 

[00:46:20] Sharon Gee: You know, our e-commerce platforms who are not participating and figuring out how to build to the new protocol is gonna be in trouble. Yeah, they are. But that's why we're doing that as fast as, you know, hopefully faster than our competition is. 

[00:46:32] Scott Messer: And then what about like on the data and identity side where. 

[00:46:35] Sharon Gee: Oh. 

[00:46:37] Scott Messer: I'll just leave that as the question, how does, how does the data and identity like ecosystem change with this new channel? Or do we think of it as the same way of like when Google shopping took traffic, like how do we think about what happens to data identity, the graph providers, the onboarders. 

[00:46:58] Sharon Gee: I think that once they become both very, [00:47:00] very important, but also under deep threat, right? Because let's, let's think about where, where that discussion happens. So let's, let's imagine that you are a fraud provider and you've got a really good understanding and you're already of, of where that experience of that, like, let's do a fraud call during a checkout happens. 

[00:47:16] Sharon Gee: Right now it's been happening on the dot coms. And so these fraud providers have like deep integrations to e-commerce platforms, right? But if the checkout experience has now just changed. In these agent channels, and the person who's doing that fraud call is actually a payment provider with a wallet. 

[00:47:32] Sharon Gee: Sounds like they better go build integrations into those companies really fast. Right? Or build partnerships where, Hey, we're already integrated here. You're already integrated there. Let's make sure that data can flow cl cleanly. So I think 

[00:47:42] Scott Messer: see. Similar volume of fraud check or anti-fraud checks? Um, I think they're not a fraud provider. They're probably an anti-fraud provider, but I'm, I'm with you. Uh, the, so same volume of anti-fraud checks [00:48:00] potentially, but fewer individual buyers of technology. So whereas you might have lease you, you might have a million websites to sell your service to, you might be down to like a hundred. 

[00:48:13] Scott Messer: Of these agentic commerce checkout platforms, and certainly if it gets rolled up into a payment provider, now you're looking at like six. 

[00:48:24] Sharon Gee: Yeah, well, or like, like we said, websites aren't going away, so there will still be volume through websites, so there, but what you're saying is there might be less to eat right on those, on those sites. So I think certainly 

[00:48:33] Scott Messer: less to eat. And then when you wanna assign, when you're saying like, oh, agentic is gonna explode the volume of commerce that's gonna happen in the world, the truth is that there isn't room for a thousand providers to service 10 million customers. It's, there's gonna be room for like fewer providers to service. 

[00:48:52] Scott Messer: But might only be 10 people who need to actually buy that service. 

[00:48:57] Sharon Gee: Maybe I think that's a little more doom and gloom that I would [00:49:00] get to because I think that assumes an adoption rate that we're not yet seeing. Um, as you say, I think that there's gonna be a lot, like, at the same rate that we are identifying changes happening in the tectonic plates of who owns that shopper experience, moving towards age agent. 

[00:49:14] Sharon Gee: What we also know is that there's a lot of products that can't be surfaced and bought with a thumbprint on the age agent surface. And so tho that cohort right. Products where maybe the answers you have won't be answered directly within that experience. It's just that the checkouts are in the same way that you had dot coms and you had marketplaces and everyone's like, oh, with Amazon websites are going away. 

[00:49:35] Sharon Gee: That didn't happen, did it? It didn't. 

[00:49:37] Scott Messer: Some parts of them did. Okay. I'm, I'm totally with you and I, I have one, I think I have one more question that I, I just have to ask your opinion if you're providing this, uh, product information discovery things, it's a, you know, more than just the color, the shape, the size, um. Where do you think the, the chatbot [00:50:00] LLM, like starts and stops with doing its own research versus just reading and trusting what's in the file that's provided in the feed? 

[00:50:10] Scott Messer: Like what if they're like, you know, actually I don't think that that's a good dress for Italy in the summer because X, Y, Z, and it starts disobeying the file. You know, like, is this a concern or am I just super out on a limb here? 

[00:50:24] Sharon Gee: No, no, it's a super valid concern. Think the answer is, um, if those channels are trying to sell you ads, which they are, um, they're gonna need to get it right. And I think that, um, we've seen them be, you know, a, each channel has its own tuning of how, like how much are they summarizing that information that they're scraping off of the internet and what are the sources they're scraping it from, right? 

[00:50:44] Sharon Gee: Like if you say, Hey, I wanna know about this brand and the brand's providing information, which is an authoritative source on product data, generally speaking. And you've got data that you're scraping from websites, not just the brand's website, but on. Social experiences or coming from Reddit or Quora or you [00:51:00] know, like the world's opinion about that brand. 

[00:51:03] Tom Limongello: I haven't heard that in a while. Okay, 

[00:51:04] Sharon Gee: I never thought that Quora was gonna be a, a reputable source of site of like sighted data in 2026, and yet here we are, it's often showing up as referenceable site, right? So what I think is really interesting is that marketers now need to have this multichannel approach to how they're validating the social proof of those products, whether it's reviews. 

[00:51:24] Sharon Gee: There's, you know, there's brand provided information there and then there's consumer validated information and then there's like it, where else does it live? And so, 

[00:51:31] Scott Messer: of hallucinated information. 

[00:51:34] Sharon Gee: yeah. And so I think the, you're gonna say, oh no, brands don't have control of that. No they don't. And you know what? They haven't had control over, over that on search for a long time. 

[00:51:41] Sharon Gee: It's gonna just get, that's just gonna amp it up. And so it becomes even more important for them to have like an AEO / GEO strategy that's multi-channel, not just their own.com. 

[00:51:50] Scott Messer: Yeah, this has been fascinating. Thank you very much. 

[00:51:55] Sharon Gee: I love, 

[00:51:58] Tom Limongello: But, um, you really [00:52:00] perform. So thank you and thank you for joining. 

[00:52:02] Sharon Gee: I appreciate that. Thanks for the time guys.  

[00:52:05] ​