[00:00:00] Brian: Welcome to e—

[00:00:02] Phillip: Welcome, Palm Springs, to Future Commerce, the podcast at the intersection of culture and commerce. I'm Phillip.

[00:00:07] Brian: I'm Brian.

[00:00:08] Phillip: And we are closing out eTail.

[00:00:11] Brian: It is late. Four hours long.

[00:00:13] Phillip: Yeah, we have been here. We came snow, came hail, came— come what— what's it?

[00:00:22] Brian: For you. For you. Come sleet, come winter storm.

[00:00:26] Phillip: What is the name of this winter storm? Hernando. Hernando 2026.

[00:00:29] Brian: Oh my gosh.

[00:00:31] Phillip: No, but we made it. And this has been— navigating the retail ecosystem has been the theme of eTail Palm Springs. We're gonna cover a little bit about the sessions that I sat through— a bunch of briefings, spoke to a ton of companies—

[00:00:44] Brian: You did. You got so much chat in, so many chats in.

[00:00:48] Phillip: This is probably the most reporting that I've done ever at one show, and really kind of hit it hard.

[00:00:56] Brian: We haven't even talked about it yet. I want to hear all about this. Maybe I should do a little interview with you. I got some little nuggets myself, but let's— who did you talk to?

[00:01:05] Phillip: Oh my goodness. So this has been, like, a packed week for me. Obviously, eTail this year— it's Palm Springs every year— and it kicked off, I believe, Sunday night. And so we're recording this on Wednesday. I'm probably going to wind up talking about this for the next week or so, so check out— man, whenever this lands, you're gonna have to check out The Senses, where we'll probably have a little bit of a recap, and then we'll do one or two pieces on Future Commerce Insiders where I think we're gonna talk about specific takeaways— which we'll get into. But I met with a ton of folks. Met with the folks who were in charge of the agentic product over at Bluecore. I spoke to our friend Ryan Murden, who is over at Brandfuel. Love Ryan.

[00:02:01] Brian: Ryan's a legend, man.

[00:02:02] Phillip: I have some takeaways this week— some themes that you guys won't even believe, things that you just won't even believe. And then I met with CommerceIQ; they're doing some really interesting agentic stuff. Met with Decile, met with Pavan Pant, who is from Extuitive— and they're a new company. I'll talk a little bit about some of these new early entrants into AI creative for media buyers. I also met with Modem, who's, I think, a later-stage version of that, much more mature— I think they're four or five years old in their journey. Resolve AI I met with. Attentive— so I met with Keri, who's the CMO over there at Attentive, and the list goes on and on. Had a ton of meetings, and then I caught a ton of content on the main stage. I closed out the week with our friend Lauren Livak Gilbert, who's over at the Digital Shelf Institute. Mhmm.

[00:03:08] Brian: Part of the— which one?

[00:03:15] Phillip: I won't even try, Brian. No. So— Digital Shelf Institute— and they're Salsify. There it is. That's the one. We had met— actually, we did this guest appearance, Lauren and I. Brian, you'll have to listen to it later, because—

[00:03:32] Brian: I heard it was raucous.

[00:03:34] Phillip: It was bananas. We had a great time. We had already met, and we commiserated over on another podcast. We were on the Always Off Brand podcast with Scott Osman, and that's where I saw Sarah Levinger for the first time. Amazing. You know, it's funny— I had followed her for years on Twitter, but Sarah is an expert in consumer psychology and behavioral strategy, and that's what her whole company does. It's called Tether Insights, and I've followed her for many years, but we had never actually spoken before, and never met before.

[00:04:13] Brian: She's provided lots of feedback on content over the years.

[00:04:16] Phillip: She's been such a pleasure, as just, like, an online acquaintance, and very gracious with her time. But now we met for the first time. We had a great sit-down on the podcast. But I closed out the eTail with a killer panel. Lauren did a great job. I've never seen a five-person panel so efficiently governed.

[00:04:39] Brian: I haven't seen very many panels very well governed.

[00:04:45] Phillip: This was "Mastering the Omnichannel Customer Journey from First Click to Lasting Brand Loyalty," which is a lot of stuff to say. But it was Mandy Herrmann, who's over at Allivet; it was the head of marketing at Histoires de Parfums; it was the CMO of HalloweenCostumes.com.

[00:05:10] Brian: Yeah, fun.

[00:05:11] Phillip: Yeah, at Fun.com. Yeah. That was funny. I know that you knew that.

[00:05:21] Brian: Was it— was it Di Lyngholm? I've had Di. [unclear]. Oh, okay, cool. Yeah, I've had Di from Fun on one of my panels at eTail Boston.

[00:05:30] Phillip: They're really good. They had really awesome insights. But Matt Eisner, who is, like, a global alliances director over at Algolia— which I have already written a piece this week about my feelings about, you know, sort of sponsored panels at events. So you know how I feel about that. But he had something really interesting to say—

[00:05:59] Brian: Yeah. Your point was, like, "Let's put up some people that have interesting things to say. Let's put them on the panels." So, he had something interesting.

[00:06:08] Phillip: Well, he did. He's got a really— he just basically had this thought process around— when originally Algolia had this mode around "search and merch," and that's kind of what they've been known for for a long time. Doing search and merch in the old days, they would read UTM parameters, they'd know what you were searching for and where you came from, and so they could personalize the page when you landed there. With ChatGPT and other agentic search, they don't know much of that when you're arriving at the page. From Algolia's perspective, they're getting people much further up the funnel when they're landing them in the search-and-merch experience. They're creating more adaptive experiences, as people are now refining their behaviors based on what the answer engine brought them to. And this has basically made them rethink how the product delivers experience.

[00:07:29] Brian: So basically, the customer can no longer be Cotton-Eyed Joe, because you don't have to ask the question, "Where did you come from, and where did—"

[00:07:38] Phillip: "—where did you go?" That is the greatest thing you've ever said, hands down.

[00:07:43] Brian: I've had a few of those this trip.

[00:07:44] Phillip: By the way, this is why we need a three-person podcast— because other people would laugh at that, and I would just admire the joke.

[00:07:52] Brian: I can laugh at my own joke. How's that?

[00:07:54] Phillip: It's pretty good. I mean, there's so much more to say, but that's sort of my experience. We'll cover more of it. But how about you? We've been dividing and conquering the whole of the show.

[00:08:05] Brian: Yeah. Yeah. So I got to catch a little bit of content. I also ended up on more panels than I expected.

[00:08:12] Phillip: Yeah, you did. You actually did a lot of content, because I wasn't even able to come—

[00:08:16] Brian: —and sit in on them. Sorry about that. No, you were on the plane. So, on Monday, I had a great panel on clienteling— which, I actually— they gave me some choices, and I picked clienteling, because I love clienteling, and I think it's how everybody should think about their customers. And so I had a really diverse panel. This was really cool. I had Corey from Lowe's, Bianca from Wella, and Tim McKeough from Ernesta. And it was really interesting— like, think about those three different brands. Like, Lowe's has a very diverse set of customers. Everything from people who have no idea what they're doing— they're walking in like, "I don't even know what my project should be." Totally. "I want to do something, but I don't know what—"

[00:09:10] Phillip: "—it is about."

[00:09:12] Brian: "But I know what it is." Yeah. Up through extremely well-spec'd projects for pro builders and prosumers. Absolutely.

[00:09:24] Phillip: Probably the widest range of customers that most people have to serve at all.

[00:09:28] Brian: Oh, I agree. I agree. From where they're at in the process of buying— it doesn't get much bigger than home improvement in terms of width of potential spot in the process. And then— it's really interesting, because Wella doesn't have stores, they rely very heavily on, like, stylists. And it's a family of hair-care brands. Yeah. And so their way of clienteling is actually working with the stylist, who then works with the end consumer.

[00:10:08] Phillip: I'm sorry. I guess I'm not following. What is the name of the brand? Wella?

[00:10:13] Brian: W-A-L-L-A.

[00:10:15] Phillip: Okay. Sorry. That's the one.

[00:10:16] Brian: Yeah. They have a number of smaller brands, including Good Hair Day— which I'll get into in a second— oh, yeah— because I ended up running a panel with Lindsey from Good Hair Day.

[00:10:34] Phillip: Good Hair Day is the— so how do we know them? Have they been around?

[00:10:39] Brian: ghd, yeah. They've been around for a while, like thirty years or something.

[00:10:42] Phillip: I feel like we've had them at something before.

[00:10:44] Brian: Oh, you probably— I don't think we've had Lindsey, but I think that we've had people from either Wella or Good Hair Day involved in some stuff before.

[00:10:52] Phillip: Okay, go ahead.

[00:10:53] Brian: Anyway— so, yeah, they're a lot less direct connection to customer. They definitely are trying to build that, but it's mostly through these stylists. And so they have to kind of work with archetypes in many ways, and build trust, and trust their stylists— the stylists they build relationships with— to do a good job with their products and show them off. And so it's almost like B2B, like, practically.

[00:11:25] Phillip: I see. Well, yeah, there's a professional line, like a— who else has this? It's like—

[00:11:33] Brian: Well, you know, a lot of other hair care, you know.

[00:11:36] Phillip: For sure. Yeah, like aestheticians, and— there's, like, lots of professional brands that have it—

[00:11:41] Brian: —or very similar to our friends at Kravet, Jesse.

[00:11:46] Phillip: Yeah, that's where I was gonna go to— interiors. Totally. Home— you have a professional line, but you also sell to consumers.

[00:11:52] Brian: What's interesting about Good Hair Day is that they're selling the same product to both. It's the same product.

[00:11:58] Phillip: It's not a prosumer version of the product.

[00:12:01] Brian: No. Right. Yeah. Versus, like, Dyson, which has a lot more variation in their lines.

[00:12:09] Phillip: For sure, right? You have the industrial version, or you have the— right. By the way, it's Wella. It's W-E-L-L-A. That's why I couldn't find that.

[00:12:16] Brian: Oh, sorry. Did I say "Walla"? I meant to say "Wella."

[00:12:18] Phillip: That's okay. No, I was just, like— it wasn't clicking for me.

[00:12:25] Brian: And then Ernesta has custom rugs. Yeah. And they have a whole bunch of showrooms, which are really cool. You can't actually buy the rug in the showroom, though, because it's fully custom. And so they have a really interesting customer journey, from, like, initial— maybe they see something on Instagram, or they're researching a project for their house, they need to, like, do a little floor covering— to getting into the store and touching it, because these are not cheap rugs. I mean—

[00:12:56] Phillip: No, I'm looking at it. Yeah.

[00:12:56] Brian: It's, like, $8,000 for a rug— which, I know, isn't even at the top of the rug spectrum, but it's not, you know, a $200 Costco rug, that's for sure.

[00:13:10] Phillip: There are cheaper options. The Paula is five-foot-by-five-foot, which— I don't know, that's a very small rug— for $600.

[00:13:15] Brian: Yeah, yeah, yeah. So you can say they're kind of mid, and I think that you get a lot of quality for that mid range. But people want to come in and feel the rugs, and they don't understand what they're going to do to the room. And so the point of transaction is not in the store, but the decision point can often happen in the store.

[00:13:36] Phillip: Yeah.

[00:13:37] Brian: And so being able to build those relationships— also, people don't replace rugs very often. They also don't replace either— hair irons, often— a straightening iron. They were saying that they keep them for ten years. They're quality products. And so there's a lot of questions about what does clienteling mean in a very, very long repurchase cycle, and with minimal cross-sell as well?

[00:14:11] Phillip: You know what's funny— there was another session about this. It was our friend who I would say is probably the most discerning purchaser of technology of anyone that we've ever had on the show.

[00:14:24] Brian: Who's that?

[00:14:28] Phillip: Kristen Flor Perret.

[00:14:31] Brian: Oh, Kristen's a beast.

[00:14:31] Phillip: VP of Brand at Sur La Table. We've had her on the show before. We'll have to make a note to link up our prior experience with her on the podcast in the show notes. The moderator of the panel was Jaysen Gillespie from RTB House.

[00:14:49] Brian: Oh, yeah.

[00:14:49] Phillip: But she was on a panel with Goop, Stitch Fix, and Chinese Laundry— who is also— Sarah Zurell. She's here?

[00:14:59] Brian: She's here. What? Yeah. Oh, I want to see her. I'll have to message her.

[00:15:03] Phillip: I was sitting with a couple of folks; I really enjoyed that panel. Kristen had a very similar conversation around— every technology vendor in this room wants you to buy their technology that's going to recommend that you buy another pot when you just bought a pot. But your pot is supposed to last you ten to fifteen years, right? You don't need another pot. No, right? You need things that go along with it, and that's where—

[00:15:35] Brian: It better last me ten to fifteen years.

[00:15:37] Phillip: Yeah, for the amount of money that you're about to spend on it, right?

[00:15:39] Brian: Right.

[00:15:39] Phillip: And so it's the "buy once, cry once," right? Yeah. It's the technologies that have been built—

[00:15:50] Brian: —and "measure once, cook twice." That doesn't work.

[00:15:55] Phillip: You're in rare form.

[00:15:55] Brian: I know. You missed one of my puns that snuck in there earlier too.

[00:16:00] Phillip: There's just such a challenge in the technology space here— it is the out-of-box nature obviously doesn't fit every model. And in these cases where it's not, like, a replenishment, right— or a "if you liked this, you'll also like the exact same thing in a different color"— that might work for shoes or jeans or a top, but it doesn't work for durable goods, right? Where a lot of these folks are— if you're buying technology today and it's a considered purchase, you're probably buying technology because you're trying to increase the efficacy of your paid performance media, performance marketing, because you're spending a lot to acquire a customer.

[00:16:43] Brian: 100%, 100%. In fact—

[00:16:45] Phillip: You want to convert them better.

[00:16:46] Brian: These companies have to go get new customers. That is actually the challenge. They're not going to resell them. I mean—

[00:16:53] Phillip: AI is just going to do all the work. That's what I keep hearing. We should just be talking about that.

[00:16:58] Brian: Because that was every follow-up. You know what we can do— we could literally just say "AI" back and forth to each other the rest of the pod, because that was basically it.

[00:17:06] Phillip: But I do want to talk about that, because you said "go get new customers," and that was the focus of one of the sessions and one of my briefings— which was, "we need better customers." We are acquiring the wrong customers, and we know that sometimes— like, we make trade-offs.

[00:17:24] Brian: That's right.

[00:17:24] Phillip: So if you are acquiring the wrong customer, then you're spending money to acquire— you're spending money in vain— you're acquiring the wrong customer, and that customer is not going to be valuable for you. And what if AI could predict that customer for you?

[00:17:42] Brian: Oh my gosh.

[00:17:42] Phillip: And here's an interesting note— a lot of the agentic conversation here has steered more towards the simulation model. So a lot of— yeah, I've talked to— there were two in particular. CommerceIQ seems to have an extraordinary and a very sophisticated solution, according to my briefing. (Hey, how's it going? Good. There's people packing up over here.) But their solution seems very mature in covering FTE-style roles, and how you would manage your operations. If managing operations of your business were relegated to an agent, and that agent was sort of taking over a role in your organization— then that agent should be working around the clock, and it should be queuing up things ahead of time, and it should have a job description, it should know what it needs to do on a weekly basis. And so— "Hey, you need to produce a report on Monday morning for all of your brands that are in your brand portfolio, for your media reports."

[00:19:00] Brian: "If you don't, you'll have a performance review. You'll be under one—"

[00:19:03] Phillip: All of those things— but it needs to be also in the brand template that you usually do. All of these things are what humans would normally do. You may be paying an agency to do it— 75 reports that come in every Monday. Well, you usually only do it for 75 of your top-performing brands, because those are the ones that you have account management for. But now you have an agent that does it for you. You can do it for all 320 brands in your portfolio. So this total shift of thinking is, okay, now how do we take that to the next level— which is, what are the insights? And it starts to go into, "Well, it can scenario-plan." So now I'm not sitting with dead time thinking, "How do I pull on the AI?" and think to myself, "Okay, what are all the questions I should be asking of it?" And then I go, and it thinks for a whole long time, and now I've got a lot of downtime waiting for its response. Instead, it's coming to me proactively, because it's working when I'm not working. And it's coming to me proactively because its role is made to look for problems. The marketing agent is looking for a segmentation issue.

[00:20:15] Brian: Mhmm.

[00:20:15] Phillip: And your segmentation issue is showing, "Hey, I have a segment that is an invaluable segment." It's not a valuable segment for me as a marketer to continue to invest in. Right? Because I keep buying these customers that have high CAC and low LTV. Those are things that you would have to, as an organization, surface, invest in, create segments of your own accord, then look at, create a dashboard around, and then bother to look at, and then make a decision around— all the time.

[00:20:49] Brian: The speed to do all that is usually weeks, because—

[00:20:53] Phillip: Weeks, if not months. And then to communicate that change to the organization— all of those kinds of things are just dramatic. They don't give you agility.

[00:21:?] Brian: Right.

[00:21:?] Phillip: Okay. So, if CommerceIQ is to be believed, it's a very big shift of the way that you would work.

[00:21:?] Brian: I'd more say that's at the top end of the scale— especially if you're in retail media, or if you're managing multichannel retail. I think that— to be— that is a sophisticated agentic commerce org. When we keep hearing about agentic commerce, I think everybody keeps really dumbing it down to "agents are going to do shopping for you." This feels more like agentic—

[00:21:42] Phillip: —commerce. Yeah.

[00:21:42] Brian: Totally, to me. I think the "agents doing shopping for you" thing is still percolating, whereas this feels like something's going to happen in the not-so-distant future. Like, it's already happening in some ways. And some of these tools— you know, there's a lot of them that are all saying— what I'm kind of taking away is, context is what AI is capable of doing for you, like getting context. Because what used to happen is, you'd have all of these data sources in all these different platforms, and you'd have to figure out how to pull them out, figure out what the relationship between the data was, and understand what that meant for the bigger picture. And now, because AI can sort of absorb all of these different data sources and have context, it does what a person can do. It just does it a lot faster, because it has all this context up front. So I felt like the word "context" was used a million times.

[00:22:51] Phillip: A lot. Yeah. A lot. Why don't we move over there?

[00:22:54] Brian: Okay. We're going to do a little walk-and-talk. We're going—

[00:22:56] Phillip: We're going to do a little walk-and-talk, because they're packing up. The hazards of trying to capture the show at the very end of the show. What are some of the other points of note? You spend a lot of time talking to people in the hallway track, and I think the hallway is where some of the most valuable insights are happening. Absolutely. You are also getting a POV; they're talking about what's going on in their organization. I have a lot to say about this, but what are you hearing?

[00:23:26] Brian: So the thing I was hearing was that analytics is the thing that people are really focused on. There felt like there was a lot of new tools focused on assessing data. And that's where investment dollars are going, because people can see the benefits. So there's a lot of new entrants into that market that are going to put pressure on Triple Whale and Northbeam. Right. And maybe they're additive to some of those, like, tools. But a lot— I think that there's gonna be— and I've seen some demos— tools that will replace them that have AI built in across more channels. And there's a lot of talk around, like, the in-tool AI not having enough context to actually be helpful, and then using stock data.

[00:24:32] Phillip: Synthetic data.

[00:24:33] Brian: Synthetic, and also, like, stock. Like, there was just— not actually based on anything. It's just, like, whatever was set in this tool before they installed it. Like— yeah— and so—

[00:24:48] Phillip: That doesn't even make sense to me.

[00:24:50] Brian: I know. I know. So— what I heard a ton was, a lot of the data that AI is basing decisions on right now, in-tool, is almost all wrong. And it's not capturing the big picture. And it's that if you're just going to use native tools with AI, you're not going to have a good outcome— and everyone pretty much universally agreed with that.

[00:25:17] Phillip: Wow. That's so interesting, because I'm hearing a lot of people who are really excited about— well, it wasn't, sort of, AI dashboards. It was more about ways to integrate to different parts of their stack.

[00:25:39] Brian: That's exactly it. So that's what everyone's saying— is, we need to have the holistic viewpoint of the data across all of our different data sources, and a lot of the in-tool AIs were not really accounting for all of that other data.

[00:25:58] Phillip: I cannot tell you how many times CRM has come up in this conference for me. Multiple people have basically said CRM has become a talking point in the last two years. Head of CRM is definitely a big role— Head of CRM is an emergent role— especially in the mid-market. I see people who are spending a lot of money on building out that. It's no longer "marketing automation," it's no longer just "CDP." CRM is, I think, the de facto standard. It's like a role-based hire now. But the next level here is— I'm hearing more and more people talking about go-to-market tooling. So, where people were talking two years ago about CRM, I'm hearing many people— and not just people on the SaaS side, where you would think— like, people aligned to revenue and sales or partnerships— that are talking about go-to-market tooling. I had a conversation with at least two different merchants— one was aligned to B2B, but one's not aligned to B2B— who are using tools like Clay. I heard about it— specifically, that's meant for go-to-market, but for prospecting and enrichment, list enrichment— for finding things like influencers, finding creators and new partners to go work with, people to go meet when they're trying to open doors to a vendor. Everyone, in their role, they're trying to open doors, they're trying to figure out who do you call at this organization to start a new strategic partnership. Like, "I want to create a collab." This is the new economy. And what might have been a go-to-market tool in the past— prospecting— is actually being used by people in this space. And I'm going to write more about this in earnest, but go-to-market tooling has come up in five conversations that I had in the last two days. That is wild to me.

[00:28:32] Brian: Yeah, it is. And it's like— I think your point about CRM is important, because CRM is sort of like the foundation and, like, data holder for what happens post-go-to-market-tooling, and even pre- sometimes. So I think you're absolutely right about that being a continued focus— like, basically, who are people? How do you get in touch with them? What do you need to say to them? Yeah. And I think there's a lot of emphasis on that. And also a lot of entrants on the service side as well, because AI is making leaps and bounds in terms of, like, being able to handle requests from people, to answer questions that it could take a salesperson— or, I'm sorry, a customer service representative— like a whole week or two to go find the information around. And so I see this as actually kind of the flip side of what you're talking about, which is— "service is the new storefront," to go back to 2022. Yeah, exactly. I think that's actually kind of coming around full circle now that the tooling's in place to make it happen— native checkout, things like that. So if you put go-to-market plus an AI-enabled service department that's actually selling— like, what does this mean? It means that the internet doesn't become a set of asynchronous activities. It becomes real-time relationship management.

[00:30:26] Phillip: And— you know, it's funny too, because there's other tooling right now that's basically doing the same thing, in specialized verticals. If you're in the creator-economy space— I leave from here tomorrow morning, I go to SoCom, which is the social-commerce conference in LA.

[00:30:49] Brian: It's going to be good.

[00:30:50] Phillip: It's going to be really good. My panel is not about this, but I met with Modem. And Modem— their whole solution is— well, they have a very large platform that does a lot of creator management, but their solution is the storefront that powers Sephora storefronts for creators. But they're powering Macy's storefronts for creators. They're powering JCPenney. They're doing all these different brands. So if you're a creator and you have a storefront on some retailer's website, Modem probably powers that. One of the key responsibilities in that creator management is— well, you've got to source creators, but you also have to manage all of their— there's a whole dashboard of analytics and data that you have to manage. But there's also the affiliate-channel management, the affiliate payments— and you don't wanna make them sign up for another affiliate program, so they're probably on, like, an existing one or something. Like, there's all kinds of— this is its own tooling and its own ecosystem unto itself. It is its own form of CRM. It's, like, its own custom form of CRM. And so, that, to me— that's why it's commerce, it's ecommerce, but it's like a specialized— social commerce has emerged to be its own thing. It requires its own tooling, and it's a specialized vertical. I asked— you know, I had this conversation with Wendy Wildflower, who is the founder. And she was saying that AI is sort of the core in the platform— it's handling a tremendous amount of some of the orchestration, and they're using AI to develop it, but they're not pushing AI on creators to use AI within it. Right. But, anyway— I don't wanna—

[00:32:59] Brian: No, I do know. I think you're onto something— which is, basically, like, a lot of internet usage is just social media, right? It's like— development management is, like, how people actively use the internet now. And what does it mean for that? What's the next evolution of that? And I think this is actually all related— like, how do you connect people together that need to be connected together? I think one of the problems we have, though— and something that I don't know who makes this— but, like, consumers and even business people are gonna need some kind of tooling to, sort of— I don't want to say "protect themselves," but, like, manage this themselves. Like, there needs to be a receiving technology for a lot of this, like, outbound engagement. Because if the internet becomes one-to-one relationships, then there needs to be some sort of tooling set up to allow the receiver of all of this to be able to manage all of it, because it's gonna get overwhelming.

[00:34:27] Phillip: Well, that's— I think—

[00:34:29] Brian: It's not just your email platform—

[00:34:35] Phillip: Well, that's— isn't that where AI comes in?

[00:34:35] Brian: You don't manage your business platform— you know, you have all these tools to, like, basically try to get after someone's inbox. And, like, I feel like the inbox has, like, four layers of management—

[00:34:50] Phillip: You know what other people are saying— I spoke with Ometria— Josh from Ometria, who was at our event. We had a great event at eTail.

[00:35:03] Brian: World of Future Commerce had a great party.

[00:35:06] Phillip: We did, yeah. I mean, eTail came alongside of us, and we did the official after-party. We did this Tuesday night. Yeah, Tuesday night. We did this great poolside event, and did it in conjunction with our friends who helped to co-sponsor the event.

[00:35:27] Brian: Virtuous, RapidCanvas, Ometria, and Bidnomic.

[00:35:31] Phillip: Yeah. And so thank you to everybody for coming out. That was a great showing of our community, so thanks, everybody. We literally turned some people away. There was—

[00:35:41] Brian: —like 200-some people on the wait list.

[00:35:44] Phillip: Yeah, it was a really killer event, and the food was amazing.

[00:35:48] Brian: Oh, yeah.

[00:35:49] Phillip: We had a great meal spread. Gosh, they are very, very good. We should— just kidding. Man, we can't go back there, because we have another dinner tonight, but I want to go back there.

[00:36:00] Brian: I know. I would eat food there every night.

[00:36:03] Phillip: I want to go and get the real spread, because— anyway. But I was talking to Josh from Ometria, and he was saying they had an RFP that came in six days ago. He had to work late last night to finish, because it's due today. I said— I was like, "Oh, yeah? Really?" He's like, "Yeah, I've got 160-some questions left." And I was like, "In the age of AI, you're worried about an RFP?" He's like, "Oh, give me a break, man." He said, "This was over 700 questions. Without AI, without Claude— Cowork, in particular, which only came out a month ago— we never even would have taken this."

[00:36:49] Brian: Right, totally.

[00:36:50] Phillip: "We would never have even taken it."

[00:36:52] Brian: Yes.

[00:36:52] Phillip: Right? But he said it's Claude Cowork.

[00:36:58] Brian: Yes, Claude Cowork.

[00:37:00] Phillip: It allows him to have access to all of— you know, his desktop, his desktop tools, his access to his email and to his calendar— like, it was the orchestrator that pulls everything.

[00:37:13] Brian: And I'm assuming that, as they go respond to this using AI as a tool to answer 700 questions—

[00:37:19] Phillip: By the way, I don't know if this is, like, proprietary information, so we might have to check, but—

[00:37:24] Brian: I mean, we're not saying any names, so it should be okay.

[00:37:27] Phillip: I'm sure it's fine. I don't know.

[00:37:29] Brian: The thing is, the job of the RFP responder is the same as the code developer. They become a shepherd and a reviewer rather than a writer.

[00:37:46] Phillip: That's exactly true. It always has been.

[00:37:47] Brian: Yeah, kind of. Yeah, I mean, you and I have responded to RFPs—

[00:37:52] Phillip: Oh my word. So many times. Like, my whole life, for a stretch of a few years—

[00:37:56] Brian: It was insane. And you know what? Those RFPs— I think we're getting the wrap-up signals— those RFPs are deserving of an AI response, because they might as well have been generated by AI.

[00:38:10] Phillip: That's true, yeah. The 700-question RFP— I have a really hard time believing that anyone actually reads, in earnest, the 700 responses.

[00:38:23] Brian: I agree with that completely.

[00:38:24] Phillip: Especially when they're getting fifteen, twenty businesses to bid.

[00:38:29] Brian: This is box-checking at its finest. Yeah, for sure.

[00:38:33] Phillip: And they're having AI read the responses.

[00:38:36] Brian: Yes, they must. Yeah, totally.

[00:38:37] Phillip: At that scale.

[00:38:38] Brian: What are we doing? We're making documents for machines to— we're having machines write documents for machines, of course. We don't know this, but— I mean, maybe it's conjecture for this particular instance, but—

[00:38:55] Phillip: I've seen a lot of machine-generated documents. There was one company back in the day who used to proctor all of the RFPs. And it was a nightmare. I would sit there and I would just say to myself, "This is just unfit." Yeah. I couldn't live with myself, because commerce is just— it's unfit, you know.

[00:39:24] Brian: That's it— commerce is not "FS"?

[00:39:28] Phillip: No. "For F's sake." Yeah. Anyway— so, I just said— you know, there'd be hundreds and hundreds and hundreds of questions in, like, the most locked-down XLS file that you've ever seen. And it seems that nothing's changed in the last decade.

[00:39:46] Brian: It's actually just gotten worse. Yeah. It's gotten worse.

[00:39:50] Phillip: Yeah. Because now— it's not like you get the same Excel file— it used to be one company that used to proctor all of it. Now anybody can generate their own cockamamie RFP.

[00:40:01] Brian: Yeah. Totally.

[00:40:02] Phillip: And it can be hundreds and hundreds of questions.

[00:40:04] Brian: It's, like, an order of magnitude larger. It's probably better than what we got before, truth be told. Because— look at the— like, think about it. You could have an AI go assess your current tech stack and needs— that's true— and crawl it and be like, "Hey, here are the things that I need." And then you could be like, "Hey, here are the things that I want."

[00:40:22] Phillip: By the way, that sounds like such a panacea. It will never, ever work that way— because I cannot tell you how often I tell AI to go to a website, grab something for me, and it is extraordinarily obvious to me that it never went to the link. It's true. I say, "Hey, write a podcast description for me based on the transcript at this link."

[00:40:46] Brian: Yeah.

[00:40:46] Phillip: Okay? And the transcript's there, and it makes up some random mumbo-jumbo, and I'm like, "You just made that up based on the title of the page, didn't you?" Right. It's like, "Oh yeah, sorry about that." And I was like, "Shouldn't you just follow the link and go visit the link?" Yeah. "I'm sorry. You have every right to be upset. I should probably go visit that link now." And I'm like— yeah— it is very human. It's as lazy as a human.

[00:41:12] Brian: It's a junior. Like, this is your intern. Someone put it a way recently that I really like— and I forget who it was, so I'm very sorry to whoever I'm stealing from— but it's a drunk college intern.

[00:41:26] Phillip: Yeah. So— anyway, my big sort of pet peeve is, is it really going to accurately crawl and assess your site, and figure out all the gaps for you? I don't know. I don't think so.

[00:41:41] Brian: I know. There could be some sort of— this is the whole thing— it's like, these are the wrappers. These are the wrappers that people are building, that's like, "Okay—"

[00:41:51] Phillip: They aren't gonna be worth much.

[00:41:52] Brian: They're not gonna be worth much, because they are just features of AI. But it's like— okay, you do need a wrapper that basically forces the AI to, like, go review every page.

[00:42:03] Phillip: Here's— okay, just— speaking of "wrap," let's wrap this by saying— last year, a lot of the eTail conversations felt like a lot of fluff. Like, "Here's what's coming," or here's what's—

[00:42:20] Brian: You can say "projecting."

[00:42:21] Phillip: Sure, why not? Yeah. But I felt like it was a lot of merchants who were talking about "Here's what we were doing with AI," and it was basically ChatGPT. "Use ChatGPT." It's true. And this year, I hear a lot of vendors actually saying, "And here's how our agents are doing actual work." And I saw it demonstrated, I see what they're doing. And, you know what, a lot of it might be the unbelievable advancement in the frontier models in the last year that are enabling and unlocking these things.

[00:42:56] Brian: Actually, right? Totally. In fact, one hallway-track convo I had— and I'm not going to say who it was, because I'm not sure if they meant this to be private or not, but I don't think they did— their company will hear a feature request from a client in a video call. They will input that video call into ChatGPT and say, "Create a document that's a spec for this, that I can review, for this feature." Then they'll take it and they'll put it into Claude Code. And they'll have Claude Code actually go create the feature. They'll have their CTO do a review on the code, and they will have the feature live in a matter of days. That is the future—

[00:43:40] Phillip: That's the pace of change. That reminds me— I'm in a Slack group put together by Fireteam.

[00:43:55] Brian: Oh, yeah. Yeah.

[00:43:56] Phillip: Jess Bachman—

[00:43:58] Brian: He's great.

[00:44:00] Phillip: Jess Bachman over there. And he's doing some extraordinary stuff with AI. And they have a lot of agents they've created with n8n and just orchestration. It's basically like a Zapier for everything. And what they've done with it is— it does lots of things, but a lot of things originate from their internal client calls. So on their internal client calls, or their internal creative meetings where they'll be talking about client work— so, every few days or every week, they go through every client account, they talk about all the work that needs to be done on each client account, and they're on a Zoom, and they're just talking like they would speak to Alexa: "Here's the things that we're going to do." So they've got, sort of, a wake word, and they say, "Here's what we're going to do." Then it has the context from the whole of the conversation of what all they're going to do— the action. So then they tell it, "Okay, now, everything you just heard, this is what we want you to do with it." And then every time that they have an internal meeting, every Zoom call from that internal weekly meeting gets ingested by this workflow, and then it breaks up, agentically, all of these actions into multiple briefs. The briefs get dumped into the client folders, which another work agent picks up, and that turns it into actual content. So then it writes the script, then that turns it into the next thing— which is, like— it starts pulling from— you know, it starts creating actual work content.

[00:45:48] Brian: This is so wild. This is so wild.

[00:45:50] Phillip: They tee it all up.

[00:45:51] Brian: It's so wild, man. This is what I mean— is why people are worried about— everyone— I keep hearing people be like, "Oh, yeah, you can do more with your same team." Right. That's such BS.

[00:46:04] Phillip: Mhmm.

[00:46:05] Brian: This is actual job eliminations.

[00:46:07] Phillip: I don't think it is. I don't think it is— not in the traditional way. Nobody is going to have this happen and be like, "You don't need to hire anyone."

[00:46:12] Brian: The pathway for people to come up—

[00:46:20] Phillip: You do. You need to hire people that know how to make this.

[00:46:23] Brian: Well, yeah— except for, those people are— I mean, how many people do you need to actually know how to do that? Like, compared to the amount of people that were doing manual work before.

[00:46:36] Phillip: Brian, let me say it this way. Yeah. The people that are, like, 10-person or five-person agencies that know how to do this now— yeah— will be 40-person agencies next year, and there'll be 400-person agencies in three years.

[00:46:46] Brian: Sure.

[00:46:47] Phillip: And they're the ones who are at the forefront, and they're cutting the new paths. They will have to either build these skills or hire people that have these skills. They're not getting rid of any jobs.

[00:46:56] Brian: They're not getting rid of any jobs. I agree with that.

[00:47:00] Phillip: The other people are going to— they're all doing just fine, and they're all gonna service all the legacy accounts who don't know what they're doing anyway.

[00:47:09] Brian: I don't know, man. I think that— maybe we'll need time— but I actually believe this is job destruction. And I'm not against it. I think that there'll be more—

[00:47:22] Phillip: It's transformation. Yeah, it's transformation. Things change because— you know, how many people are in the offices changing copier toner these days?

[00:47:33] Brian: Something I heard at Manifest— and this is—

[00:47:36] Phillip: People were barking about job destruction when the internet came around, and said that fax machines and Xeroxes are going to go away.

[00:47:40] Brian: No— and I'm not saying I'm against it. I think there are a lot of jobs that are probably not, number one, worth doing. Right. But one question I did hear at Manifest, and I think it's worth asking— not Manifest? Yes, not Manifest. Oh, okay. Yeah— which is: what do we lose when we eliminate the mundane?

[00:48:03] Phillip: No, I agree.

[00:48:04] Brian: Yeah. And so I think that there's things that— is it context? Is it a sharing of skills? Is it knowing what the outliers are? Is it judgment about making decisions— which is the work now, those are the work. And so if you don't have the process to, like, learn how to get judgment, then you can't make decisions that are good. Or you just export them to a machine.

[00:48:37] Phillip: Obviously nobody has any of the answers here, but I find it really interesting. I'll close with this— my last word, and then I'll give you the last word. I thought this was super interesting. I had— who was it? Let's see, I've got to pull it up here while I'm saying it; I should have had it chambered when I was bringing it up. It was a panel that was hosted by the Attentive CMO, Keri McGhee, with Cozy Earth's CMO, George Davis, and the VP of Commerce from HEYDUDE and Crocs— his name's Tommy Kowalski. And Keri asked— which, by the way, I've never seen anybody as an exec at a technology company deliver an actual, like, journalistic— like, a full-on, like, media-quality, journalistic-quality interview. She killed it.

[00:49:50] Brian: I wonder if she had ChatGPT help her.

[00:49:54] Phillip: I asked her. I told her she did a phenomenal job, because I did a briefing with her— and, actually, her background is in media. She was at NBCUniversal.

[00:50:?] Brian: Oh, got it, got it.

[00:50:10] Phillip: And then after that, she spent eight years at Zillow helping them build their B2B business.

[00:50:12] Brian: Oh, I wonder if my friend knows her.

[00:50:12] Phillip: But she asked— I think she asked George Davis. She said, "What do you think about all this AI stuff? Where do you think this is all going? Are you excited about this? Are you hopeful?" And he said— he's like, "I'm half excited, I'm half afraid." She goes, "Well, why are you scared?" He goes, "It's taking all of our jobs. No software vendors left. No one to sponsor this conference next year—"

[00:50:52] Brian: "—to take me golfing."

[00:50:57] Phillip: And then he said, in the next six months, right now AI is supplementing their workflows, but in the next six months it's going to own entire workflows without any human intervention.

[00:51:09] Brian: Yes. That's what I'm talking about.

[00:51:11] Phillip: Yeah. His doomsday scenario is like— you know, "Maybe next year we'll have an eTail, but question mark on the year afterwards." And I am not so sure about— I don't think that that's true.

[00:51:22] Brian: Yeah. I mean, there's a bunch of doomer articles that just came out recently. I was, you know— it's like, come on, we'll figure it out. People are needed. There's lots to do still. In fact, most people in this room are overworked. That's the truth.

[00:51:39] Phillip: Yeah.

[00:51:40] Brian: Like, the mundane things kind of kill them, actually.

[00:51:46] Phillip: Oh, that's true.

[00:51:47] Brian: So— I— RFP responses. Yeah. Oh, man. Actually, it may have made it work worse, because now poor Josh feels like he has the tooling to respond to a 700-question RFP—

[00:51:59] Phillip: To do more of them.

[00:51:59] Brian: Yeah. A 700-question RFP in six days. That's crazy.

[00:52:04] Phillip: Yeah, that is crazy.

[00:52:05] Brian: Yeah. You know— we know that's the reality. It's—

[00:52:07] Phillip: It's only accelerating from here. Last word?

[00:52:09] Brian: Right? Last word is— I read it in a Dirt piece. That's Daisy Alioto's publication. And, unfortunately, I'm forgetting who the author of the piece was, and I should go look this up. "Consolidation is power." And I do believe that a lot of this consolidating of information, consolidating of, like, tasks, and consolidating of systems is a form of power. And so, you know, maybe the better you get at consolidating things, the more power you have as a business. And so— something to think through.

[00:52:52] Phillip: I like that. All right. Well, you can find more episodes of this podcast at futurecommerce.com. Go check out all of our eTail content, where we'll have all of our recaps, insights, briefings, and a little bit of our overview and deep dives— that'll all be at futurecommerce.com as well. If you're not subscribed to our newsletter, you can get that over at The Senses, at futurecommerce.com/subscribe. We also have our weekly insights briefing called Insiders, and that's also at futurecommerce.com— shocker. We'll have more from eTail. We've got Shoptalk coming up in just about a month's time, and we are doing an after-party at Shoptalk as well. If you want to get on the list, we do have limited space. Go to futurecommerce.com/events and get on the list, because the sooner you sign up, the quicker that we'll be able to evaluate everybody to figure out who can be in the room where it happens.

[00:53:46] Brian: Again, we had, like, a 200-plus-person waitlist for our eTail party, so get on the list fast.

[00:53:53] Phillip: It's very important. If you made it to the end of the show, we want you to be in the room. futurecommerce.com/events. Thank you so much to eTail for bringing us out.

[00:54:05] Brian: That was great.

[00:54:05] Phillip: We're gonna be at B2B Online. We'll also be at eTail East in Boston later this year, so lots more to come with our friends at eTail. But thanks for listening to this episode of Future Commerce, and that's it. Commerce shapes the future, because commerce is culture. We'll see you next time.