[DECODED] Three-Party Commerce: Trust in the Age of Agents

Phillip: And welcome back to Decoded, a podcast by Future Commerce, presented in partnership this season by our friends at Commerce. I'm Philip.

Lindsay: And I'm Lindsay.

Phillip: And today we are in episode three of five. And this is called Three Party Commerce: Trust in the Age of Agents. Here's a paradox I want to put in front of all of you listeners and viewers out there that should keep every brand leader up at night: 27% of millennials in our study, called New Modes, is now trusting an AI platform more than humans for product recommendations. And the study was very expansive. We surveyed 1,000 consumers and asked them, what are the things that are keeping you up at night with regard to shopping? You can get a copy of that study right now. Want you to go grab it. It's at futurecommerce.com/decoded. But they share their deepest insecurities with ChatGPT, and they ask Perplexity about sensitive health conditions. And they let AI shopping agents and assistants into the most intimate quarters of their consumption patterns. But those same consumers will abandon shopping carts the moment that you ask them to create an account. And this duality is what is at the crux of every brand leader's positioning, their strategy, and the solutions that we're trying to provide and to solve and put into place right now. So Lindsay: Commerce –– y'all are at the forefront of integrating AI capabilities in the merchant experience. What are you seeing, today before we get into this episode, as merchants are grappling with this new AI mediated relationship with their consumers? What are you seeing?

Lindsay: So something that we're seeing a lot of right now is that our customers are having to shift their mindset from channel management to model management, which is just kind of this whole new era of the internet that we're in. So historically we're optimizing for the channel, but the channel is no longer the focus right now. We have to optimize for this more adaptive relationship with an AI that is going to know different things about our customers and about the entire landscape. And we have to figure out how to be interpreted through those models that we don't fully control as brands. So I think that the old game was distribution. The new game is intelligibility. And brands that win are gonna be the ones that understand the model and understand how they can adapt their message to the new modes of interacting with consumers.

Phillip: That's very well said, and it perfectly sets us up for our next conversation. And today, I can't wait. Our guest is Sharon Gee, and we're diving deep into the psychological barriers behind this AI trust wall and the implications of agent to commerce and what brands need to understand about answer engine optimization in this new reality. Sharon, without any further ado, welcome to Decoded.

Sharon Gee: Thanks for having me.

Phillip: Lindsay, now three episodes into this, and we are, I think, discovering that there's more to this AI boom, this AI uptick. And we're hearing a lot, Sharon, about the data in this report and how other people have been synthesizing this data. I want to hear a little bit about your perspective. And I want to kick us off a little bit from your role over there at Commerce. Tell us a little bit about what you do over there.

Sharon Gee: Yeah, sure. My role, I'm the Senior Vice President of Product for our AI solutions across the Commerce family of brands and products. That means the AI solutions that live within the BigCommerce platform, Feedonomics, and MakeSwift. I lead the product team that builds and delivers and supports those solutions. I've had an opportunity to be at Commerce now for six years, originally joined to lead the relationships with the channel partners that enterprise merchants need in order to grow. Think Google, Amazon, Meta, TikTok, and now these new emerging channels like OpenAI and others, Perplexity and the like. It's been really interesting to see the evolution of how the data and the merchant of record settings that are required now have to be externalized in this new agentic era. So that's a really fun problem space to be in as we help merchants. We work with 30% of the internet retailer 1,000 send their catalog data to these channels so that they can hopefully be considered and more discoverable people and the eyeballs all turn towards a more agentic search or LLM based consumer AI kind of search methodology and mode.

Phillip: That's actually what we're going to dig into here is that in our research together, we found that a quarter of Gen Z and millennial consumers now trust AI recommendations more than human ones. The data is available, we'll link it into the guest guide of how we arrived at this in surveying consumers. But we found that this is the arrival of retail's first potentially post-human interface. It's a demand for hyper-personalized shopping experiences. And so this episode is going to dive into how LLMs are becoming an intimate commerce companion. So with that in mind, because a quarter of these Gen Z and millennial cohorts are trusting AI recommendations so intimately, what's happening psychologically, Sharon, when these people are starting to trust them over, say, recommendations of their friends?

Sharon Gee: Well, I think what's interesting is that trust comes with lots of sources, right? So one of the things about an experience is, do I trust you is based on, are you an expert? And so one of the things that we're finding is if my friend whose opinion I don't respect about fashion, for example, tells me what I should wear, I probably am not going to trust their recommendations. However, if I now have an answer engine LLM that sits in my pocket that knows absolutely every single piece of information about my favorite brand's catalog, where they manufacture their products, what mill they make their wool in, right? Notes from the designer around what features were designed into this product that I want to be inspired to buy. I think the narratives we tell ourselves are important and I think that when we talk about psychologically what gives us trust, it's did I have a good experience? And in this case, the experience we're having is I'm telling an answer engine that I'm going on a trip, I'm going to my friend's wedding in Italy, it's going to be really hot, I want to find a dress to wear, it needs to be in a size ten, it needs to be under $300 It used to be we would sort and filter based on price or size and it's like three words in a search box and now it's like our whole lives get poured out to these answer engines because it allows them to have the context they need to give us a hyper personalized answer, whether that query is focused on help, I want to go on a vacation during the December time frame and I need to leave on the twenty sixth and be back on the first, what is available to me in my budget price range somewhere warm? Or if it's, hey, I'm looking for a new tweed jacket from Buck Mason, right? Like I think these are the kinds of things that we need to consider.

Sharon Gee: When we think about what creates a trusted experience, it's, is this the product I want? Is it available near me? How much will it cost? And do you know enough about it to be able to help me make my decisions? And those decision making criteria increasingly are all about the context around how humans make decisions like what are the key features of this product? If you're thinking about technical specifications, you're buying consumer electronic equipment as an example, there's all these technical specs that I can definitely trust a source if the answer engine is sourcing all of their research for me and are able to contextualize that against the entire world universe's information set around recommendations and consumer validation of that product. So I think if we break it down into the data is the foundation for the trust and the merchants that participate and understand that are going to be able to take really serious advantage in this early days of being able to help answer the query by providing the data that allows an answer engine to make a good recommendation that allows me to have that trust and personalization as a consumer.

Lindsay: I love that. I kind of want to take it a layer deeper. So the research shows that 46% of millennials and 41% of Gen Z use these platforms daily for things like product research, but also for things like emotional support. And Sam Altman has said people share the most personal shit of their lives with ChatGPT. So taking that into account, these AI platforms are becoming intimate companions and not just tools. What does that do to how we interact even as merchants or just as humans with this new system and technology?

Sharon Gee: Well, take everything from a lens of what kind of opportunities does that present to our customers. We're in the business of helping brands be discoverable and growing online, right? Whether that's on their own experiences by helping them build a beautiful, compelling branded storefront or whether that's helping them get their catalog data visible everywhere their shoppers are. So when I think about what what that consumer behavior shift means, to me, it means you need to be able to meet the consumer's query where they are, but then also have a deep respect for how precious that query and data is. So for example, we really want to understand what consumers are saying in the black box. We want to know. But a question about a product might also be combined with medical information, or it might be combined with intention to buy. And so when we're working with we work with a lot of the largest channel partners out there. And as we're working with them to say, hey, you don't have an analytics API yet, but our merchants really wanna understand what kind of data can they give to you in order to create a really powerful experience that meets a consumer's desires and expectation for a personalized shopping experience, there needs to be this understanding that all of that data is kind of put together. And so it's really about how do you externalize the data that you know that you have that can answer a query in that context because if you're an you know, if you're asking questions as a consumer about the you know, the deepest, most emotional parts of our lives and and oftentimes the products that you buy associated with that.

Sharon Gee: Right? I've got neurodiverse kids and they're freaking amazing, but I the the level of, like, mom hurt that goes into figuring out what kind of products can help what kind of sensory blankets or what kind of, like, sensory swing could I buy for my daughter. Like, these are, like, emotional decisions that have shopping intent. And so really, it's about how as a merchant can you externalize the data that you need in order to help someone who's seeking you, right, in many cases to solve a problem or to be inspired or to for whatever reasons we all shop a sense of identity, but also to find really quality products. How can you externalize the data that you have in a way that allows that conversation to be very natural with an LLM and the LLM knows enough about you to say your name and know that this consumer query resonates with what it is that you're looking for, things like safety details or otherwise. So we move kind of into this structured and unstructured data action mind space that we need to understand, which is how can we give signals to these new channels around, yes, I have the products you're looking for. Here's how to contextualize your understanding of them so that when somebody asks you a question, you say my brand's name, as an example.

Phillip: Well, I think that's such an important distinction there to draw is that the the merchant has two big jobs to accomplish. Right? One is: we are still creating branded storefronts –– as much as I hear a lot of people saying websites are dead, very much not dead. Right? They also have to empower data that goes out to all these channel partners of which one an emergent giant channel partner is shoppable answer engine as a channel. And that's, think, is this thing that everybody sees it as an emergent opportunity. And I think everybody wants this arbitrage opportunity, where it's like, it's this thing that's going to print money in the future, apparently, eventually, potentially. That's what we want. Right? So, so but I think it always Sharon comes down to the quality of your data. And it always comes down to the availability of that data. And I think that's the that is the thing that separates the the desire of that traffic or the desire of the discovery ability in the in the business versus the actual the actuality of it coming through as a as a strategy in the business and executing on the strategy of the business. And I think what never gets connected in these kinds of conversations is actually how you get to the brass tacks of getting that data exposed in the organization. Right? So like, how do I get clean data? How do I get it connected? How do I get it set up in my organization so that I can get to that place where it is? I have a website, but what happens beyond that? And I think so, so many people are powering data and all of these channel partners, and they're doing it with maybe they're using feedonomics, maybe they're using this. But how do they do something beyond that? I guess that's the question that they're probably asking is that is AEO really the answer? How did they make that happen? Right?

Sharon Gee: I think it has to be both. Right? So as as you say, well is AnswerInjin is going to to to kill the website? No. But it's going to significantly change how it needs to to behave. So we've got a few truths we kind of, like, say to ourselves that we're finding as the new axioms in this kind of AI era. Number one, data is the storefront. It's like a it's a refrain for us. Right? It's it's about how do you externalize that product data in an API or in a MCP or in a brand agent in a way where your brand agent is going to be able to enter the chat when ChatGPT or Perplexity or or Google are looking for answers. Right? Because now we've moved out of this era where it used to be a shopper and a merchant. Right? So Sharon would go to Google and a merchant had bought it out on Google. They plopped me on a .com. I sort. I filter. I search. I'm looking for products and they have that first party data. They track my behavior across the website, and then they try and get their conversion rate to go from two to 5%. Right? That that's the old era, shopper to merchant.

Sharon Gee: We've now entered this era where there's a agent in the middle. So now it's a three party world where you've got a shopper, an agent on behalf of the shopper, so a shopper's agent like OpenAI or like Google and like Gemini, and that is it's acting on behalf of the shopper interfacing with the merchant. Well, very soon, in the new protocols that you're seeing emerge like a to a and others, there will be a merchant or a brand agent. Right? So you'll have a shopper, a shopper agent, a brand agent, and a brand, and all of them will be in the chat at various different systems because now you have you know, in various different interaction points because now I, Sharon, send my agent to go do all of my research for me so I don't have to have 700 tabs open for the dress that I want to wear to my friend's wedding. I can have a summarized view of that. My agent did the summarization for me, but wouldn't that be so much more powerful if that agent could interact with an agent that knows all about the Veronica Beard new 2025 season, right, as

Phillip: An example.

Sharon Gee: So these are examples of what how does how does the data need to show up? And it's we think about data's the storefront. It has to be externalized. It has to be accessible. That's where the new protocols come in like MCP, which is like a fancy wrapper around an existing API essentially. It allows agents to take action on data and perform actions within merchant systems as well, like querying information, asking questions of the catalog. And then you've got this other truth. So first one, data is the storefront. Second one, consumer is the channel. This interaction is going happen anywhere somebody can interact with a chat window, right? So whether it's going to be an agent experience within our favorite consumer apps that all of us are using or whether it's going to be within WhatsApp where we're interacting with a brand agent, wherever the consumer is, whether they're scrolling through social, searching on Google, on Amazon, whether they're typing into OpenAI ChatGPT, your data has to be where they are. And then I think the third one that's the most interesting around AEO versus getting data to the channels, you need to do both and it's because agents are customers too.

Phillip: Right.

Sharon Gee: That's the third truth. It's that if 55% of our traffic is now robots essentially interfacing on the internet, looking for information, wanting to reference information, doing research, contextualizing information so that they can serve up these contexts in new browsers or in new conversations, you need to build your website in a way that not only acts and is optimized for humans, but is also optimized for agents. And so when we think about that, that means there's new interaction paradigms that we need to be considering. How does your website serve that data up when the audience is a known robot without eyes as opposed to a human who likes to look at HTML and see pictures and have a really consolidated view of that data? So these are kind of the three truths that we think about when we ponder how does a brand or a merchant participate in this new agentic world. It's like, let's make sure that we're making decisions that consider the new truths of data is the storefront, the customers, the channel, and agents are customers too.

Phillip: That actually supports what we found in that there's likely a a correlation to a generational adoption of AI too because in the distillation of our research, we found that Gen Z unsurprisingly uses AI for product research 11 times more than, say, boomers do. So And when you're talking about AI fluency becoming a baseline, we're also talking about there's certain types of products, there's certain brands that service a certain consumer that will be quick to adopt, there will be certain ones that will be certain modalities of shopping that will be quick to adopt that will service that consumer and meet them where they are, because that's what the channel that become channel of nativity for them.

Phillip: It makes perfect sense to me then that your platforms there are serving them in this way where you have multiple, like you were saying before, there's this evolution of the consumers research, the consumer shopping journey, and it's happening at this new point of interaction. Gen Z, though, I think is the only it's the lowest band on our ability to research. I'm also seeing for whatever it's worth, in my own kids behavior and Gen Alpha that they're also in there shopping as well. We just can't ask them about it. But that's anecdotal. Anything that you you could add to that is that sort of the generational divide. And I would even say that while we were talking about consumer behavior here, too, I do think that there are generational divides in the operator side of the business when we're talking about the way that we implement the tactics and strategies to here, because I think that that has a lot to do with the way how fast we move in organizations to in leadership.

Sharon Gee: Yeah. I think we're seeing it too in behavior. Right? So it's not like all of sudden the channel showed up and said, hey. We all wanna build a conversational chatbot. Those have existed for a long time. It's that it's that consumers want to search using words. We want to speak in our native fluency in whatever language we speak, and we want our answers to come back to us as easily as we consume them. Like, I absolutely believe we are the last last mouse and keyboard generation, like Mhmm. The ones we're working right now because this interaction model of native conversational semantic fluency where you can just speak it and it comes back to you, or you can just snap a picture of your empty living room and say, design me a new living room in this style, and now show me all of the products that are available on X, Y, Z merchant website that match the aesthetic that I want. That's such a better interaction model than sitting on a computer, like, tethered where I didn't get to take a picture and have it magic and design and co-collaborate and co-create with me. So I think what's really interesting is the people who get the people who understand natively what is available. Like ten years ago, it was like the social generation, right, when all those new social agencies emerged and it was like, hey, there's this whole new generation of people who interact in this channel and this is native to them.

Sharon Gee: We're just seeing the same thing with a native AI experience, I think. Um, and I think it but what's really interesting to me is how fast it's not just the new tech the new generation that is adopting it because a lot of you I know, actually think that this is one of the biggest opportunities for B2B brands to kind of, like, leapfrog some of their B2C experiences because so much of business that is complicated was really tricky when you have to have all of these various different business processes. And one of the things that's like really amazing about LLM and AI technology is that it can simplify a lot of that capability by putting together lots of different steps and consolidating that and kind of closing down all of the necessary process steps into something that's much more streamlined. So I actually think that brand agents made available on B2B websites is gonna be, like, one of the most amazing things. As an example, instead of just, like, finding t shirts and headphones on an answer engine like ChatGPT, what if I was looking for a replacement part for a Caterpillar skid steer? And I could take a picture of the part and it could just show it to me and it already knows me and because of OFF, it knows what I have like the products that I have previously bought and it is able to show me my pricing for that as a whole as a Mhmm.

Sharon Gee: As a B2B buyer of that brand. Like, to me, that's such a huge opportunity for the B2B world because they are the experts on the products, and they know and have the authentication and the customer books and this customer specific pricing and availability that I actually think it's not AI is not just powerful only in the hands of the biggest companies. It is a democratized technology that can be implemented in companies of all sizes. And I think that's one of the most exciting things about the future of commerce is it finally got fun again. Instead of like a bunch of the little guys optimizing data to the big guys, There's this big democratizing event where the seismic shift allows those who understand that their consumers best to really participate actively in this new world with these new consumer kind of interaction paradigms. Totally.

Lindsay: I love that. I think something that is interesting to me from the data that kind of relates to this is that the data shows that 27% of millennials trust AI more than humans for product recommendations, and yet 63% will abandon carts when forced to create an account. So there's this real disconnect between the trust in a brand versus the trust in AI. And I'm really curious to understand your thoughts on that, Sharon. Why? And how do we navigate that when we still have to exist as brands and with some traditional ways of interacting with consumers with this convergence of AI into into the space?

Sharon Gee: Well, I don't think consumers are dumb. Right? We're all consumers. I'm willing to trade information as long as I get something out of it. People are willing to give you your email if you give them a 30% off discount code. Like, that's a thing that happens. Right? So I think it's all about where in the journey does the trust happen. We know we are not going to just create an account in order to maybe interface with someone where we feel like we should get that information before having to make that kind of a data exchange commitment. However, imagine how seamless that could be, and I think this is why you see the players like you know, we've announced recently some really interesting relationships with folks like PayPal and other channel partners, like, who are who are very much focused on how the auth and identity component of this already works. Because if you are already a known entity set aside all of the complications around things like fraud, which we can totally talk about for a long time. But when you know who the shopper is and you know what they want and you know you have history on what they've purchased before, you can start to do really amazing personalized things with them. And so that trust is really being built, as you say, within that shopper experience where when I ask a question, I get a good answer. And so the more of those positive interactions you have, right, you see in the research of the PDF that we were talking about, kind of there's an awful lot of trust in brands like Amazon, right, where you've had so many good interactions where you look for something, you find it, it's delivered to your doorstep with a photo of that product on your doorstep the next day.

Sharon Gee: Those are the kinds of things that build trust. So when we think about what kind of interactions need to happen, how will you as a brand be present in these kind of trusted experiences? When we think about the new protocols that are being developed by folks like Google, like and AP2 and others, I can imagine very near term this kind of construct of a conference call. Right? You're interfacing with Gemini, and Gemini says, hey. You're looking for that specific product. You know, I've got I've got a brand agent over here who knows all about the new releases. Why don't I add them to the chat? And then all of a sudden oh, I'm interfacing with my assistant who's brought in their assistant to like be a source or an expert on x y z. I would be much more compelled if Google then said, great. Would you like to auth into your account, or would you like to click a button, use your thumbprint, and automatically create an account so that I can give you better pricing or make a pricing estimation just for you or estimate how long it would take for that product to be delivered to your doorstep. These are the kind of experiences that brands are going need to be able to participate in. And of course, that means they need to choose partners that know how to and have built to those protocols in order to be able to support their participation in that next era. So I think that's how you build trust, understand what kind of consumer experiences you're trying to unlock and then make sure you show up with relevance with the right partners.

Phillip: I mean, how I'm not sure about either of you after, I don't know, three years of using ChatGPT and knowing literally everything about me and now bringing up things, I think, you know completely I think in in a very good way, I think unexpectedly saying, oh, well, when we're talking about this thing, it's like, oh, I well, you're a runner. You should also consider this. It's starting to make those sort of like out of pocket recommendations to me, things that I wouldn't have thought of. I'm starting to depend on it in a really heavy way. You know, you talk about auth. They've also just rolled a sign in with OpenAI account. I'm like, it's one click away for me for being like, it's the gateway for everything on the internet. You know, as a shopping gateway and destination to I think from an off perspective we're the friction, I think there for millennials, I think is probably if I had to guess what is the what is the point it's like, as a millennial or an elder millennial, having maintained accounts all over the internet, I think that was the issue. Like we just parts of our data is strewn everywhere. One hand to shake, I think is probably just better. And maybe this helps to mitigate that.

Sharon Gee: Well, I think you'll see that happen already. Right? Auth is a fairly solved problem, and so that's why you see the players, like, who doesn't already, in many cases, have either a PayPal, a Venmo, or a Google account, right, or an Amazon account, or a Meta account. Any one of those can kind of, like, help solve this problem, and I think the people who have hardware signals, so the phone holders of the world, right

Phillip: Totally.

Sharon Gee: The people who have authenticated account status. So think about the Googles and Microsofts where their account disposition or their off disposition of knowing who you are. You've got payment companies who have, like, the concept of a wallet. Anybody who has the concept of a wallet is in a very good a is very good position because they're much further along in terms of knowing who you are and identity and authenticity is critical in delivering these kinds of experiences. Right? And so I think when we look at that list of who the channel partners are who are going to be empowered to create these kinds of experiences, in many cases, it's going to be who has authentication data and trust already. Right? Like, I already trust Google with my email and with all of with with all of my device storage and all of my photos and everything else. Right? So I already trust Amazon with all of my credit cards that have ever bought something from Amazon. Uh, so I think when we think about who are the players in this space who are gonna have a really strong position, it's the folks who have authentication data, who have a good construct of the world's internet. So that's why you start to see some of these AI companies releasing new browsers for the first time in a really long time.

Phillip: Yeah. That's an interesting Yep. There's the

Sharon Gee: Construct around how are people searching on the internet? What does query look like? What are people looking for? Does intent data look like? And then those who have good signal around the world's catalogs, whether that catalog is of the avail if you're in the hospitality or in the travel space, that catalog is what is the available inventory for things like booking airlines, airfare, or hotels. In our category for commerce and retail, it's obviously product catalog availability and inventory availability. So I think how do you participate in this ecosystem knowing which part of the data pipeline you are to deliver that powerful experience via AI becomes really important. Then it also kind of helps you know which partners you should be looking inevitably for developments from because they have a pretty strong leg up in terms of being able to deliver an experience that's really personalized.

Phillip: I guess just to come back to that point real quick, because I don't think I was so eloquent in getting there, it's that we're creating a new set. You've talked about all of these different ways of establishing trust in the way that we've built sort of the modern internet. Talked about this transactional trust, right? We have this transactional trust through this identity layer of trust. There's this new psychological interior intimate level of trust that we have with this new this other this other relationship that's developed. It's quite new. It's three years old, right? We have we've it has a little bit of like sense of like some of our relationships, and maybe like my kids homework in there. And like, I don't know, there's this whole other thing. And now, what it also is, is it's a gateway to some other shopping partners. And I think that it has because of what we've talked about, and this has been talked about ad nauseam on plenty of other content and topics and shows, what it hasn't been linked up to is data around the this psychographic behavior. What I just think is really interesting there is that we can link that back to behavior around the most interior and most intimate parts of our lives. And that's something that I don't think that we would willingly give up to shopping partners. And so when when we're talking about that as a part of function of identity removing friction, I think it's a really interesting one. And it's something to, I think, keep an eye on. I think it also Go ahead.

Sharon Gee: That concept, right? Like the idea, you know that an influencer on Instagram is trying to sell you something. You know that.

Phillip: 100%.

Sharon Gee: But, there's some really interesting data that I know that you guys have kind of landed on around, yeah, but I've got trust that the AI isn't, right? Like, because it's agnostic. Look, it's citing me at sources. And I think that can be a little murky depending on how the AI is trained and how the data sources work and things like that, but also it's the opportunity that merchants have to influence that. Um, I think when you have something that is trying to provide you with a baseline understanding of, hey, here's all of this information and based on the query that you gave me, here's what I'm looking for to personalize for you, the agenda of the AI is much more clear, right? It's answer, give me a good answer so that I keep using your app. And I think as soon as we start to see more of the promoted posts and things like that, we'll get the same level of beautiful cynicism that we have around Instagram influencers trying to sell us things. But I think humans are good at sniffing that out, right? Like signal through the noise around what can I trust is something humans continue to evolve at being good at understanding? Uh, so when we think about like what kind of experiences do we need to create, what as merchants, what kind of experiences do we want to foster and offer so that we can give our shoppers a good experience and a good brand experience so they hopefully come back and buy our products again and create nice, loyal lifetime customers with high value there? I think that's one of the things that we have to consider is when what kind of information do you need to provide to an agnostic recommender of your products? And that that kind of leads you to this kind of really interesting AEO, GEO strategy that we're starting to see and emerge because it used to be the ads team would send the product data, title, description, image, color, inventory availability to Google.

Sharon Gee: And then they'd buy an ad against it, that that would get you seen. And now there's this really integrated story where the marketing team and the PR team have to, like, partner up because you'd actually need external validation from those influencers or from those moms on Reddit or from those validators of who whoever those people are answering Quora questions. Who knew that Reddit and Quora were gonna be like of the internet? But but I think that there's a really interesting different modality here where it's it's it's reviews and it's trusted sources of information. You have to put your signal out as a brand, but then you also have to make sure everybody else is putting the signal out so that the AI can look at this and say yes. Um, they say their food is good, but so do all of these reviewers on Google say their food is good. So it seems like a restaurant you might like. Right? So I think that's one of the things that we need to think about when, as a brand, you're gonna have to break down the silos between your marketing team, your PR team, your affiliate team, your ecommerce team, the marketplaces team, because all of them have the data that is going be necessary to create this kind of more integrated experience.

Phillip: Yeah. You're talking about silos. There's probably also another more traditional thing to talk through, which is you know, we had this traditional way of thinking about the funnel. I think personally we have a series that we produced two years ago here at future commerce that's part of our professional subscription, which recontextualizes the funnel. It's more of a loyalty loop. I don't think the funnel exists anyway. So but there is awareness consideration purchase. I think those things are pretty instantaneous now, especially in platforms like TikTok, where you can go from awareness to purchase within seconds. But how do you think AI is collapsing or reshaping that path to purchase? Is there a is there something wholly new here, or is this an evolution of that new loyalty loop, or or is there something different that we could examine?

Sharon Gee: I think it's category specific. Right? It's pretty easy for me to make a purchase around something that I saw like, if if I'm gonna buy a new set of earrings that cost less than $50 that I know are nickel free, that's a pretty easy consideration set versus, say, I want to buy that new velvet olive green sectional couch that has to be configured for exactly my specifications of my living room and it needs to be delivered with white glove delivery. Like this is category specific, right? So the level of research that is required and therefore the experience that needs to be surfaced within each of my surfaces needs to meet different needs, right? And then in B2B shopping experiences, there's all these additional layers of and do you know me and do you know what price I am privy to because I am your trusted customer of many, many years as an example. So I think the answer to that question is it depends. I don't think it ever is like, oh yes, for all B2C categories that are like the TikTok shop categories, we know that AI is going make it immediately possible for you to use a thumbprint and buy them. I think CPG as an example category. I have a specific kind of mascara that I like. I wear it every single day. And if I'm about to go on a business trip that I have to leave for tomorrow morning, the number one thing I'm optimizing for is can that be on my and I'm out.

Sharon Gee: The number one thing I want is can it be delivered to me by 3PM today so that I can pack it in my suitcase by tomorrow. That is not a brand consideration or like, oh, I might evaluate a totally new mascara. No, it's a ChatGPT, tell me who can deliver this product to my door by 3PM and I will thumbprint and buy that instantly if I already have the consideration, I'm already loyalty. So I think it will deeply collapse the funnel for things that you know that you want immediately. CPG is a really good example category for that. But when we think about high consideration products, furniture is a really good example. We know that consumer electronics are deeply researched. Think about everything that you'd ever buy on Adorama, right? You know, there's tons of experiences that are going to need to be able to where brand agent experiences are going to need to power significantly better or deeper research paradigms in order to help people feel confident about the decision that they make to purchase. Then, will I use my thumbprint to buy that product after I've done the research or will I want to go directly to the website? I might have more trust in the platform that has my wallet and my identity already saved where I've had successful purchases before like I do on Amazon. Like, nobody has a problem buying an expensive camera on Amazon now because we trust that that product will be delivered to us. If we have a problem with it, we we will have some sort of recourse in order to be able to address that issue.

Sharon Gee: Right? That took a while for for the internet to have us all feel safe like that, but I think that's certainly a similar kind of paradigm that we can see. So when we when we contemplate, like, the funnel and is it collapsing, I think it totally depends on the category, and I think it depends on the search and purchase more debt modality of that category. And ultimately, it's just becoming a lot more like, if you were interfacing with a human, how would they solve your problem? And so to me, that's that's a quest like, agents or customers too. How do we think about them and give them the information that they would need if they were your best personal shopper? That's how I think about the problems and the experiences we need to create for them because the intelligence is getting to a level where in some cases it's significantly more capable than humans, right, of being able to hold all that information in its mind. But then also it has this amazing ability to understand semantic search or semantic query and then be able to give us a really personalized experience. So that's how I think merchants probably need to think about the funnel is that there's no one right answer just like you were saying. And I don't think it's like there's some rule about how AI is affecting all funnels for all categories and all consideration and purchase.

Lindsay: So looking ahead at the next eighteen to twenty four months, how do you see the relationship, Sharon, between consumers, AI platforms, and brands evolving? Will brands build their own AI shopping assistance, or will they be optimized to recommend third party platforms? What do you see coming down the pipe?

Sharon Gee: Yeah. I think this goes back to that truth we mentioned, which is the customer is the channel. The answer is they will do it all. You still need to have a compelling brand experience. That's the way you show up and have people fall in love with your products. You know, if I'm scrolling through Pinterest evaluating what the best new design for my living room is going to be and all of a sudden I see that gorgeous Anthropologie home couch, when I click through, I'm gonna say, I didn't even know Anthropologie did couches. Right? And then I click through and I'm starting and I'm falling in love with all of the products that are available on their.com because I wanna understand a brand that I know and love for apparel all of a sudden has this offering around furniture. How are they addressing that? So I think that .com experience is still gonna exist. It's going to take, It's going to change shape of what kind of experience needs to be offered there. Then I also think the different modalities are going to be really critical. So merchants are going to need we're finding more and more of them are going to need to launch brand agents where they have control and they can leverage the same democratized LLM technology in order to be able to have an LLM offer a conversational semantic experience that is a shopper assistant or helps me do things that where it's an expert on the data that they have about their consumers, so that they have about my loyalty program or my loyalty status or about their product catalog that only they're the expert in.

Sharon Gee: And then of course, are going to continue to use third party LLM technology aggregating and synthesizing the world's knowledge on things like ChatGPT and Gemini. And so I think the answer is merchants are going to need to participate everywhere their shoppers are. So customers are the channel and the data is the storefront. And so what we need to be able to do is make sure that we understand at each interaction point when you show up with your brand, how is your data representing you? And whether that's a third party experience like a search, agentic search or social experience, or whether it's in a conversational WhatsApp chat that is querying that brand intelligence layer that you own via a data API interaction, or whether it's on an owned.com experience where you might have a brand agent directly that I can access on anthropology.com as an example. So I think the answer is brands are going have to do all of it, and it's a really fun moment to be in commerce. So I'm excited to see what you guys do at Future Commerce Velvet because you guys certainly have a really good view around how this has changed in the lore of commerce over time.

Phillip: Well, thank you, and thank you, Sharon, for joining us here on this season of Decoded, and thank you for partnering with us on this piece of research.

Sharon Gee: Thank you.

Phillip: Lindsay, as you all at Commerce are helping merchants optimize for an AI driven discovery, what is the most important shift in mindset that brands need to make today?

Lindsay: As merchants are optimizing for this new AI driven discovery world that we're in, one of the patterns that we're seeing is that merchants want the upside of efficiency and personalization, but they're terrified of losing the human signals that make their brand feel like their brand. So everyone is kind of experimenting in two directions at once, automating operations with AI, which is clearly producing a lot of gains, but then doubling down on authenticity in the moments that matter, including on their storefronts and making sure that the things that have always mattered around brands and loyalty still exist. So I think that's the mindset shift that our merchants are really grappling with right now is how do you do both of these well without losing the soul of your brand in this influx of new opportunities and tools that we have with AI.

Phillip: And I do think that that is the key, right? Because the brands that thrive in the Sara are the ones that understand that they're they're not necessarily controlling the narrative. They're in sort of this relationship with the customer. I think we always knew we're in a relationship with the customer, but we're now in a new era where there's a an intermediary and we're feeding systems that consumers trust. Right? And so maybe those there were other trust factors in the past. Maybe we had like consumer reports, right? Or we had like an influencer, but now we actually have an agentic system that is effectively a surrogate for the shopper. And that is where we have to aim all of our attention now. So answer engine optimization is not about gaming and algorithm. It's about ensuring that your brand shows up with authority. This has been a great episode. I'm so glad that we had this conversation with Sharon. You find more episodes of Decoded at futurecommerce.com/decoded. Get your copy of the new modes report at futurecommerce.com/decoded. And remember the future of commerce is what you make of it. Future commerce will help you to shape that future. Thanks for watching.