Agentic AI, AEO, and Retail Media: What Brands Need to Know in 2026 | Digital Shelf Insider

Speaker A: Remember, Jeff Bezos came out and said early on, your margin is my opportunity. And people didn't take him seriously on that sensor tower. Um, data that came out was like Black Friday. 30% of shoppers used Rufus and it ended up having a 60% higher conversion rate. Like that was last year. Yeah, I mean, it's only amplified now. I think the biggest thing is the agentic commerce piece, this idea at some point in time that these LLMs, um, or a personal shopping assistant or whatever it is, um, that they are going to go out and make purchases on behalf of, um, the customer. Research will say if you're in the electronic space, you have to be advertising on Best Buy and it looks like you're flushing money down the toilet. That's what it looks like. The reality is still millions of people do research on Best Buy and then they go on Amazon to buy it. So when you turn off Best Buy, you see sales drop in other places as a whole.

Speaker B: This is the digital shelf inside a podcast of Metrix Cart. Answering optimization, retail media measurement, omnichannel visibility and agent E commerce. All of these are very heavy words and there's a lot of noise out there right now in the industry. This is the playbook. This is how you work with it. This is not the thing. And it's very hard for brands to filter out what's the right thing to do right now. Today we're joined by Aaron Konan. He's the co founder and chief digital strategist at BWG Connect. He's also popularly the digital Deep dive podcast host. He's a heavyweight in the digital commerce space, having spent over a decade advising enterprise brands on everything from Amazon growth strategies and retail media optimization to navigate the massive shift into agent AI and conversational search. He's right at the center of where the digital shelf is heading next and we have a ton of of ground to cover today, so let's get started. Hi Aaron. Welcome to the Digital Chef Insider podcast. So, so thrilled to have you here because, uh, I guess I've been following you for a very, very long time. You bring in such great episodes all over digital commerce and it's, it's, it's amazing that you're here with me today.

Speaker A: Oh, super excited to be here. Thanks for the invite. Um, yeah, excited to share. This is openly, you know, I talk with 20 to 30 brands a week and uh, I get the, uh, I don't know, the privilege of kind of like consolidating all that knowledge and then resharing it out with the Network so awesome. Excited to be here.

Speaker B: Awesome. Uh, um, yeah. So I've got a mixed bag of questions. This episode doesn't have a theme in general. It's all about we dive into the retail media. Then what are the challenges E commerce brands face? Uh, uh, what are the leaders thinking right now about AEO optimizations and stuff like that? So let's dive in. M. My first question for you is around, um, you talk to a lot of brands, uh, and leaders from the space. So what is that single biggest macro anxiety or challenge that dish to shelf and E commerce leaders are bringing to you right now? What does it look like?

Speaker A: Um, I think well it kind of goes deep. We'll start off. I think the biggest thing is the agentic commerce piece. This idea at some point in time that these LLMs, um, or personal shopping assistant or whatever it is, um, that they are going to go out and make purchases on behalf of um, the customer and so what does it then look like? And for any of that to happen then they're, they're slowly breaking down. Okay, well what do I need to do internally to make sure I'm set up? And then that kind of got into one of the things that you're talking about which is probably the second biggest right now which is around aeo, SEO, geo? Um, and I just tell people like if you're not getting recommended in the LLMs, who cares about agentic commerce? Like no agent's ever going to go try to buy your product because you're not getting recommended within the agent and so or the LLM. So um, that is probably one of the biggest things that are out, that is out there. And the tough part is it's relatively new and there are no real clear answers. There's best practices right now but no single point solution. So yeah, awesome.

Speaker B: Uh, to dig a little deeper, uh, like you mentioned, there's no playbooks. Every other platform is functioning very differently. Uh, so how does the content evolve for brands, uh, in terms of thinking, for humans, for agent, E commerce, um, how does it evolve exactly?

Speaker A: Yeah, I mean I think the interesting thing there's been a couple dynamics that are going on as well which are how people are shopping is changing, um, what are they doing, how much are they reading, how much are they just looking at images? So if we look at the conversion piece and we also have to look at then showing up in search piece and those things used to be almost the same because the person doing the search was your was you know like the old Google way or the old Way On Amazon, before LLMs, how people would, would search would be black coffee maker.

Speaker B: Yeah.

Speaker A: Um, now they're searching in a different pattern, which is longer form, which is. I'm looking for a coffee maker because I, uh, that fits my, you know, modern farmhouse theme in my house.

Speaker B: Yeah.

Speaker A: This is a completely different way of searching. And so what is flipped and if we isolate it to like an Amazon and this was going to proliferate everywhere else. The old way of looking at doing your content as a whole is great imagery and then keyword stuffing essentially within the text. And so that was talking to the algorithm and then you would dump everything into the text to hopefully convert the customer, whether that's lifetime warranty or whatever it is that's built into that text. Now nobody searches coffee maker with lifetime warranty.

Speaker B: Yeah.

Speaker A: So what's fundamentally changed in these and um, search on marketplaces, we'll say is that text, people aren't reading the text, they're looking at the images to convert. And so you have to move the, converting text from that, nobody's really searching on, from the written text into, into the images, put a logo or something, put something on the image that says lifetime warranty. So when they see it, they still get the message. It helps them convert. Everybody likes to see that. But then when you're looking at the text now, the text needs to talk to not just the algorithm, but these LLMs. And so when you think about the algorithm, that used to be keyword stuffing and keyword stuffing over and over again. But these LLMs, they are not. But number one, they're not really reading the images. It takes too much energy, too much power, too much money at the end of the day to read the image. So they're looking primarily at the text. And now if your keyword stuffing, it's seeing the same thing over and over again and it doesn't care anymore. It sees it once. It's good. So now you have this revision of all this written task. So now you're touching all your visual, your images, and now you're also having to touch all of your text to now talk to, um, the LLMs, like with a Rufus, I think sensor tower, um, data that came out was like Black Friday. 30% of shoppers used Rufus and it ended up having a 60% higher conversion rate. Like that was last year. Yeah, I mean, it's only amplified now. If you look at total users, uh, even like a chat GPT, total monthly queries, total daily queries, like monthly, it's over a billion now. I Think is where it's at. So this is a new way. That thing that people are using um, routinely and they like it. So man, if you're in the content space, I think just when you thought it was slowing down a little bit, like maybe you'd optimize has all been turned on its head. Um, and if you don't. Here's one more thing is then if you don't do these things now, the LLMs, the ChatGPT or the clods or whoever, um, they, they're training right now. Yeah, right. Just they get better every day so they're training right now. If you're not adapting right now, you are not going to be recommended in these LLMs. And then does agenda it answers the next one originally like then what about agentic commerce? Well I guess it doesn't really matter. And so there's a mass scramble right now to try to update content. Um, and if you're not doing it, the competition is so. Yeah, yeah, so that was kind of long winded but it's complicated.

Speaker B: That's what it is. Yeah, that's what it is and it's very, very complex. Um, when we think for brand. So we spoke what brands got to do and what consumers the way the consumers are um, you know, searching how that has evolved. What are the retailer. Retailers exactly doing the Amazons and the Walmarts. Ah, have they kind of, you know tweak the way the rules are on for titles descriptions to adapt for these changes. I uh, mean I saw a recent update where there was an Amazon title specification changes but how much has it evolved from the end? So the brands can make these changes?

Speaker A: Yeah, I think they are. Amazon's always, I'd say a B testing um, at the end of the day they want to make sure that they're putting the product the customer is most likely to buy, not always the best possible product, but the product the customer is most likely to buy in front of them, um, and then help convert them m as quickly as possible. So those tweaks are going on but largely it's on the brands at this point in time. I mean you think about uh, from the Amazon standpoint, whether they buy your brand of kitchen spatula or another brand of kitchen spatula, Amazon doesn't care as long as they're the one that sells you the kitchen spatula they're still making money off, it's the next one up. So then it's the survival of the fittest which Amazon has pretty much always been. And then you have people who game the system and then Amazon steps in to try to correct it. Um, you know, and that's happened slowly. I guess one of the most recent ones, right, was with the ratings and reviews, uh, used to everything. If you had variated ASINs, all of those reviews would be like aggregated into one big review count and now it's split out by asin. By asin, all the variations have the appropriate one. So which, um, I understand why they did it. There's a lot of people scrambling on that front. But I guess my point is like Amazon, the tweaks they're going to make around requirements to text or images or bullet points or um, product descriptions or titles is only meant to enhance them showing up better and them related the products that they want you to buy and them being at the top of the list. So what does that look like? Well, uh, from my standpoint, when I think about it, any updates they make in that space are such that when an LLM outside of their rufus is looking and grabbing data, they're the one most likely to show up at the top as a recommendation for a product. So um, those are the tweaks that I believe that they would make to make sure they stay at the top as people still shop by far online. So.

Speaker B: Yeah, yeah, yeah. All right. Before we get back into the conversation, I just want to take a moment to talk about something that can really elevate the way you manage your e commerce channels. If you're in the consumer goods or durable space. Listen up. You all already know that staying ahead in the digital marketplace is crucial whether you're a retailer, brand or agency. But how do you get the most accurate insights in real time? How do you keep track of competitor price activity, consumer trends and constant changes in product visibility? That's where Metris Card comes in. Metris Card is a marketplace intelligence platform designed specifically for e commerce teams like yours. It gives you a holistic, data driven view of your performance across all digital shelves. Whether it's price monitoring, competitive analysis or understanding trends before they become a wave. Metricart has you covered. I've seen firsthand how this kind of tech intelligence takes brands from just surviving to thriving. If you're managing products in the consumer goods or durable space, it's a game changer. Check them out and I promise you your e commerce game won't be the same. Now let's get back to the show. I have a question on the omnichannel side. So there's a growing sentiment in the industry that chasing broad blanket omnichannel visibility can actually dilute margins. Um, you often discuss the value of strategic channel management. So how should brands really balance being everywhere with maximizing SQ level profitability?

Speaker A: Oh, I think you nailed it. I think, I think brands largely, um, and I'm not talking about the startup digitally native brands, I'm talking about the consumer brands that most of us know and use and we find in retail already. Um, I was part of this when I was on the brand side was Amazon E Commerce Omnichannel was largely relegated to the corner. Right? The crazy people in the corner trying something new and Amazon has started to take off. Um, and that was even up right till um, Covid hit. So then Covid hits and there's this mass scramble to get online. And what ended up happening was um, a lot of these companies just started to try just dumping all their products online. They were amazed. Right? It used to be if you're going to get on shelf at a retailer, like it's a long process and then at the end of it you're trying to convince them, the buyer that your product on that shelf space is going to make more than the current incumbent on that shelf space either higher volume obviously more money, at the end of the day more margin, whatever it might be. Uh, so a lot of these companies then jumped all into Amazon and they're filling out these awarded forms and Amazon's taking everything and everyone's like woohoo, look at we sold in everything. Then they quickly find out, um, that you know, and I start with Amazon but then obviously it trickles over to every other marketplace that's out there. Is that not everything is profitable, especially with the take rates they have and the, you know, people are shocked. But remember Jeff Bezos came out and said early on your margin is my opportunity and people didn't take him seriously on that. Um, and that, and that leads us to where we're at now is people are finally taking a step back and they're saying okay, what is the skew? Skew by skew, um, profitability as a whole. And that's what you have to go through now just because that skew isn't profitable, the next level people are saying okay, but how can I make this profitable? Which then, is it a kit, is it a bundle, is it a two, is it a three pack? Um, is there co marketing that can be done? And then the next level is because all this stuff is everywhere and can be sold anywhere. Obviously brand protection and all that other stuff comes in huge. Um, but what if I have different ASINs, different SKUs, different kits, different bundles, marketplace by marketplace. And then you get this true omnichannel approach, which isn't totally crazy to think, um, people use different count sizes all the time, right? Um, to differentiate between a Walmart and a Target. Well, why not do the same thing online? And so that's this evolution. But I think everybody's looking at profitability right now. I mean you mentioned it too, right? Like that's how do we make this profitable? The only way is by um, doing a unit skew by SKU profitability analysis, um, and then being willing to shut that SKU off if it's not making money.

Speaker B: Exactly. Yeah, yeah. Um, I think maybe two years ago we would have heard that, you know, um, there were brands, brand managers who would use a copy paste approach. Like say what, whatever PDP they have on Amazon, they will try either Walmart and they try to work. Does that happen anymore now? How is it?

Speaker A: I'm totally shocked that it does. I laughed about this because you do a lot in the Walmart marketplace. Everybody, uh, it's really weird, right? You go back, everybody's, you know, Walmart is a giant behemoth and it's struggling. Everybody's butting heads and now some, everybody wants, you know, a competitor, a true one to Amazon. So they're kind of rooting for Walmart.com. they lag is people will spend time, effort, money at a large scale to optimize everything for Amazon, right? And then they're like we've got it optimized. And then they jump over to Walmart and they copy and paste and they're like yeah, Walmart.com, the marketplace, it doesn't work for us. Like how much did you optimize for it? Did you spend as much time and effort and money as you did on optimizing for ah, for Amazon? They're like, no way, it's not worth it. Well, chicken to the egg here, right? Like yeah, are you sure it's not worth it? Uh, and they're like well we don't want to do that until we see results. And I'm like well maybe you don't see results until you actually do that. And that's what we see. Um, that's what we see a lot of um, which then leads to. That's right, is like these Walmart partners that are now popping up, that are exclusively focused on that piece of it. How do you get up going, how does everything get optimized? Um, we're having more and more of Those pop up. But yeah, you can't just copy and paste.

Speaker B: Yeah.

Speaker A: I mean maybe if you're a small mom and pop shop. Right. You're doing $500,000 a year on Amazon. You're making um, you know, a uh, kitchen utensil.

Speaker B: Yeah.

Speaker A: Like it's a lifestyle, it's great. But if you're a large organization, you just can't do it. You can't do it. So.

Speaker B: Yeah, makes so much of sense because uh, you spoke about SKU by SKU analysis. So I had to ask you this question. Like, you know, where, where is that? Like have brands stop doing that because it's, it's, it cannot work. Just cannot work. Yeah. Yeah.

Speaker A: Crazy. Yes.

Speaker B: Yeah. Um, we cannot, uh, you know, have this episode without talking about retail media. Retail media measurements are still very confusing for brands and uh, we're seeing RMNs move up, funnel into CTV and even bridge into in store digital signage. So how should digital leaders shift their budgets and measurement frameworks to account for this?

Speaker A: Oh wow. Um, so what I'm going to set aside is the impact of AI, um, and the LLMs on retail media because search, a lot of that is fed in on the digital side by search and feeding it. And it's just going to fundamentally change with the LLMs, um, and agentic shopping and doesn't matter if a person isn't there if it's an agent doing it. So that piece I'm going to totally set aside and we're going to focus on the here and now.

Speaker B: Yeah.

Speaker A: Um, rather than what's coming up. But anybody who's listening, just keep your eye on that ball because it's going to drastically change over the next, I mean probably in 12 to 18 months we're going to see a big shift. These uh, alums have to make money somehow outside of subscriptions. And so um, the retail media piece is probably the most um, hard nut to crack.

Speaker B: Yeah.

Speaker A: And that is because everybody's trying, struggling to understand where do I spend the next dollar and incrementality and cross channel. And we've gotten to this point now where it used to be though the, the seven touch points, you know, in marketing before somebody buys, now that's well over 40. And it's because people, when we talked about omnichannel everywhere, not only are people shopping everywhere but they're consuming and researching everywh. So it's fundamentally changed in, in how we look at it and then also the complexity of these systems and how they interact because something on, we saw that there's um, did a podcast with a lady, Megan Karun from CLAIR Data. I don't know if you know Megan. If not, I will connect you with her.

Speaker B: She's a. Yeah, I did have her on the episode, like when, you know, retail media was very, very nice, maybe like last year sometime in Jan. And yeah, it was good.

Speaker A: Oh, she's amazing, right? She's amazing. She's the only one that I found to truly break it down. And this is the thing is you can't do this on your own and most likely your agency is not going to do it. Well, your agency is biased and most people don't have a, uh, single agency that manages everything and has your best intentions. Usually people have a chunk here and a chunk there. Um, and so what she is able to do, um, is step in and do that analysis. Now, it's not cheap, right? At the end of the day, it's so hard. It can't be. You're plugging in all these different things. What people don't realize is like a lot of research will say if you're in the electronic space, you have to be advertising on Best Buy and it looks like you're flushing money down the toilet. That's what it looks like. The reality is still millions of people do research on Best Buy and then they go on Amazon to buy it. So when you turn off Best Buy, you see sales drop in other places as a whole, um, that is so complex. There's a lot of brands, like I'm talking huge brands that are, say, their data science team. And if you're at a large brand, data science team is like, oh, we can do this. They can't. They just, they can't. Um, and then they also have all these agencies who are saying they can do it. Just trust us. Like, no, you're not grading your own homework. So from the retail media side, um, you have to be able to bring in a tool, um, like a Clair Data. That, that's the only thing that I found, especially at scale, to look at all the different paid search, social retail media networks, um, because we all know all these RMS are asking for more and more money every year. Right? And I don't blame them. Um, but the, what I hear all the time is, um, from the large, even like the CPG spaces, um, I just want somebody to, to be able to prove that when I give you a dollar, you're giving me money back. And sometimes it's okay if it's just a dollar back. Sometimes it's okay if it's 50 cents back. Obviously they want more, but you know, sometimes it's okay to lose a little bit to make more money back. I was like, I say I'm like a Best Buy or something tough.

Speaker B: Yeah.

Speaker A: If you just want that knowledge.

Speaker B: Yeah, yeah. I mean, I like the fact that you share that. You know, when, if you are a consumer electronics brand and then you have an ad running up on Best Buy, you stop at the sales drop. Like that's a simple analogy, but it really matters. Like for brands, it's so important that you know they look into these aspects.

Speaker A: Yeah, I mean all cross polys, people are shopping and researching and consuming data everywhere. Um, no, no, no, not at all. Uh, what else is top of mind for you? This is fun. This is fun.

Speaker B: Totally fun. Um, so a big part of your mission at BWG Connect, uh, is helping companies connect with partners and platform selection. So the digital shelf tech stack, which is the PIM and the syndication tools and analytics platforms, uh, can get incredibly crowded and confusing for brands to choose from. So what are the kind of common mistakes you see brands make when selecting their digital tech stacks? Because it's like infrastructure, it's like the basic thing that you need to have. Right?

Speaker A: Um, actually it's, it's not talking to enough people. So this network that I kicked off, uh, nine years ago now, um, and it's still around because I was on the brand side and I'd been um, you know, trying to find an Amazon agency, I'd been burned on a web dev firm and an integrator and I just, at that point I was doing a lot of Google searches. Right. Or way back in the day, maybe you go to an IRCE or you try to find somebody. And um, what I found was I just wanted to talk to other brands, I wanted to talk to people about their experiences. So it is in this digital shelf tech stack space, but it's also on the space of, well, you know, uh, what is a real TikTok agency look like and who's working, who's not, or an Amazon agency or who can handle Amazon returns or how do I drop ship. Like there's all these things that the best way to find that right. Is by just talking to people. Right. Um, which I know from the tech stack side, people don't. I find some like people in the tech side or the newer. They don't necessarily like that. But the reality is if you're a really good business and you're doing really good things, you have a lot of good people out there that want to tell your story and how good you are. And so part of what I love to do is I'm talking, like I was saying, with 20 to 30 brands a week and just saying, like, hey, who's working great for you? Who do you love? What are your pain points? Who's helped you solve the old ones? And then I can run this network or, um, you know, I can just put that information out through, um, either, you know, my podcast, Deep Dive, or the webinars we do. We'll do 100 to 150, or the we'll do 80 to 90 small format dinners. Yeah, it's all about just being able to, once we found those great partners, just help them out and get their story out. Um, and that's. That's kind of how we built this. It's like, all free for the brand side. I just built something I wanted on the brand side. So the number one thing is just build your network and talk to people, right? Everybody in this digital space, we've all jumped into something that a lot of people were uncomfortable with. So I. People should never feel like they're asking a stupid question, because there was no. Any question your answer in the digital space or ask in the digital space, the answer wasn't there five years ago. So, you know, we're not talking about, you know, simple mathematics or something like that where we've always known the square root of four. Like, we're asking things like you're saying, like, how do you measure retail media? That's a legitimate question. And the only way you find out that brands aren't doing it on their own, they need it. You know, a tech sec partner is by talking to people. And so, um, and that's how we get connected with great people like you, right?

Speaker B: AS FRANCINE Likewise. Likewise. I, uh, mean, I would completely agree with whatever you shared because, uh, it's the same for me. I mean, I used to ask a lot of these questions to all the great leaders in the space on LinkedIn messages, and I used to get all the responses I need. And I'm like, okay, maybe this could be a conversation. Why am I just having it on this chat? A lot of people can listen to it and, you know, learn something from it. So it's nice that way.

Speaker A: Yeah, I know. Like, even on. That's what I love is, like, we could easily just. You and I had this conversation, and you didn't have to hit record on it. And then at the end of it, we were like, shoot, we should have hit record. And just literally Shared it with everybody. But that's where, like, I just love it. Um, I love people and I love E commerce and connecting them and I don't know why, but I do. I started off as a chemist by myself at the back of a lab and. And uh, then here we are chatting.

Speaker B: Um, yes.

Speaker A: What else is top of mind? Anything else?

Speaker B: I have a last question for you before we find this episode. Uh, so how can E commerce leaders separate actual needle moving AI capabilities from the shiny object AI marketing fluff when evaluating, you know, vendors, software vendors, because there's a lot of AI powered being stitched into marketing. Uh, when we look into apps currently.

Speaker A: Oh, wow. Yeah, that's a tough one because we're at the front end of it and 95% of it is garbage. 5% of it is really good. But we don't know what 95 and 5% is right now because it's so new. Um, what I'm a huge advocate for is having an, uh, what I would call an AI R&D budget. Um, and a lot of that just coming from way back in the day when I was a chemist. Um, you know, from an R and D standpoint, there's time and there's money set aside and organizations know within R and D, most of what happens with those experimentations doesn't work out. But if you do enough of those experimentations over time, you finally get the golden nugget that actually does work. And so that is the only real way to do it outside of whatever networking group, thought leadership, share groups that you're a part of so you can bounce ideas off from. Um, but right now, the interesting thing with AI and these tools are if you miss it, you don't miss it by, you know, the normal, what would be a normal six months, you miss it by a year to two years. Right?

Speaker B: Yeah.

Speaker A: Because everything in that space is moving that fast. And so, uh, again, the share groups. But have that R and D budget set aside and let your team play around with it. Now what does that mean for large organizations? Your cio, your cso? They have to be involved because if you're plugging tools into the network,

Speaker B: they're

Speaker A: going to want to know. Um, and I don't know if they set up a sandbox or whatever for you to play around in, but you have to have that budget. Um, and that includes not just tools, but other things. I've seen agencies pop up. Um, so if you don't know how to enhance Reddit.

Speaker B: Yeah.

Speaker A: Which these, these LLMs go to. Right. Retailer.com, brand.com, then Reddit. Well, how, how do I experiment with Reddit to see if it has an impact on how I'm showing up in LLMs? Unless I've got a set aside amount of money to go try it out for three to six months, maybe it doesn't work. My guess is it probably does. Um, a chat GPT agency. Right. You don't always have to be the one that's doing it and doing all the experimentations or trying all the new tech. There might be a smaller agenc there that's already doing it and they can help. Right now we see a lot of coaching.

Speaker B: Yeah.

Speaker A: Um, going on. Um, and so maybe there's an AI agency. I think we'll see those pop up outside of just your normal agency but. And at the end of the day some kind of R and D budget that allows people to go in your organization to go out and try these new tools because you can't miss it. Yeah. So that's what I would do.

Speaker B: Yeah.

Speaker A: If I was it. Chief Digital Officer somewhere. Um, oh, and hiring these thought forward people, that's huge. Yes, yes, that's huge. You can't have people who are afraid to use it. So yeah, that's it.

Speaker B: So much of science. Thank you so much Aaron for taking out the time today. It was such a wonderful conversation with you. I kind of bugged you with all sorts of questions. It's from different, different areas. It was just fun.

Speaker A: You like, you like took my whole day, uh, all these individual questions from brands and you consolidated all into like a half an hour. Let's go. That's so much fun. Well, thank you for having me on. So much fun. Uh, yeah. And uh, yeah, look forward to doing it again.