Everyone Talks Agentic Commerce, DaVinci And OpenAI Shipped It

Welcome back to The Marketing Companion. I am so excited for today's episode of The Marketing Companion, and thank you so much for all the comments and suggestions. Today we're doing one that I think is right on target. It's called Agentic Commerce and OpenClaw: The New Marketer's Playbook, and I am so excited to have Diez from DaVinci, and you're going to be so excited to hear about what they're doing because it is very new and leading edge, but also very valuable to your customers. Before we get started though, I do want to thank our two sponsors. I want to take— thank SEMrush for what they do. I use them in my own company for all the geo work we do, and also Brevo, which we use for sending out our newsletters doing so many amazing things. Check both of those out in the show notes. So without further ado, um, Diaz, I'm going to have you start out just by introducing yourself. Thank you for joining me today.

Absolutely. Well, uh, Sandy, uh, great to see you again, and thanks for inviting me. Um, so, uh, yeah, I can tell you a little bit about, uh, my background. I always like to tell people that I, uh, as an entrepreneur, third company now, but with DaVinci Commerce. I, I tend to follow data. So first was Informatica, where we're about aggregating and integrating data, and that became a big business, sold to Salesforce for $8 billion earlier this year. Uh, so that taught me follow the data. My second company was a company called SelleQuest, which got acquired by a company called Cognos to do real-time analytics on data. And, uh, And then Cognizant got acquired by IBM. And, and then now with DaVinci Commerce, you know, it's AI, which ultimately is powered by, by data, right?

That's right. And you've had such an illustrious career. I mean, we got to follow you, not just follow the money. We got to follow Diaz because Diaz is where the action's happening with so many successful exits. And I love it because you are a CEO, but you're going to be talking to marketers today. And the fact that you care so much about what marketers do, I think is a big, you know, wake-up call. And if we look at what's happening today, AI is becoming the new storefront, right? So just measuring clicks and impressions and search rankings may be measuring the wrong thing. What do you think the first thing a marketer needs to unlearn to compete in this new environment?

Yeah, that's a great question. And I think it's really a recognition that consumers have now seen a new experience and they love it. And that is the conversational experience, right? So I wanna say back in the day, although AI and LLM platforms are very new, you know, if you look at an average retailer's site and how you go and buy stuff, You gotta do a search and you end up with 500 products that all look kind of the same, and then you gotta navigate through that and click on each of them and look at the PDP content. And, you know, it's a lot of work, and you don't realize how much work it was until you actually experience what happens in LLM platforms where you just say, hey, I'm going to a party this weekend, I'm looking for a new dress and any ideas? And, you know, and then the agents go and find it for you, give you choices and so on. So I think consumers have experienced that, they love it, and marketers have now got to figure out how do they kind of look at every way by which they try to engage consumers and change it because consumers are not going back to that click-heavy experience that they, that they were, they fell in love with many years ago and now have experienced something much, much better.

Yeah. And I think, you know, every marketer, every CMO that I meet with— I just did a day-long session with CMOs, another one with CHROs as well. And, you know, a marketer spends a lot of time building up their brand voice and that visual identity. And now that you've got ChatGPT and Claude and all of this, a customer can ask a question and you get back something that was never brand approved, right? And so that's a scary situation. A lot of marketers are struggling with that. What do you think? How do marketers protect what they've built but they don't own the channel anymore? What is your advice to them?

Yeah, that's a great question. I think, you know, as much as, you know, brands and marketers thinking about, okay, all these consumers are going to ChatGPT, other LLMs, they're looking for products, we need to be there. But the first cut of being there— first of all, it's kind of interesting, right? If you go to ChatGPT, ask a question trying to find a product, whatever, you get a whole bunch of text, right? Yeah, there's a few products mentioned here and there, and if you're lucky, your brand might be one of those mentioned. But gosh, like, just a piece of text describing your brand feels terrible, right, if you're a CMO or a marketer, right? So that's sort of version 1.0 of of agent-led commerce. I think 2.0 is— I think platforms like ChatGPT have realized that consumers buy more than just seeing a piece of text or whatever. They want to, you know, experience the brand, right? And, you know, it varies by product category, obviously, but there are certain brands particularly that sell based on experience, right, and based on the lifestyle and so on and so forth. So And that's why I think recently there's been much more of a focus on experience. The notion earlier on that, oh, a consumer asks a question, you get a bunch of texts telling you here are the products with the prices and you click on and you buy, didn't work. It just did not work. I think the numbers, and Walmart was one of the ones that jumped in first and said, oh, we want to work with a fan, do that. And the numbers are dismal. It was something like 3 times worse than search, right? And, you know, and the reason is because consumers, first of all, don't just look at a product, say, "I want to click, I want to buy it," right? And that the idea that you could do instant checkout tied to just a product listing just wasn't real. And the second issue is the experience and trust, right? Experience meaning Consumers want to know more about the product. They want to learn about it for quite some time. And when I talk to retailers and brands, they keep talking a lot about the PDP pages on commerce sites. That's where the conversion happens, right? It's not the search kind of gets you there, but they look at what's on the PDP page, look at everything from reviews and ratings to experiences that actual consumers have posted there. So then the question is, if that's what gets consumers to buy, that experience is not there in the LLMs today. So I think OpenAI, realizing that, came up with this app model, which is much more experience-focused. And actually Walmart reported early numbers on that, you know, app model, and they're great. The conversions are much better. People are discovering products that they weren't even intending to buy in the first place. So I think for brand marketers now, finally, you've got a better ability to recreate that immersive experience in these LLMs than you did a few months ago.

Yeah, you know, I was just actually with Walmart at the HumanX conference, and they were talking about this power of agentic commerce and what they're doing with it. And it is really driving a lot of purchase for them from consumers. And consumers are actually happier about what they're seeing now in the LLMs because so many people were, you know, driving there. So why don't you tell us kind of what you're doing with OpenAI and Accenture? Just give us a little bit of background because a lot of people may not know about this new feature inside of OpenAI and how you're helping companies with that.

Yeah, no, uh, that's great. Actually, I was at ShopTalk a weekend ago, and I was surprised. I, I think I met about 15 or so brands and retailers, and every time I showed them this, you know, uh, product, the Agentic Brand Store, they said, wait, you can do that on ChatGPT? So yes, it's new, and a lot of both consumers and even brands don't know about it. So, uh, what OpenAI did was they created the infrastructure to develop apps, right? So think of them like iPhone apps, but it's very different in the sense it's conversational to match how consumers engage on these platforms. Now, they provided the rails of the tools, but then somebody had to actually go build the app, which on the face of it, well, okay, you build an app, but this— there's a lot that has to go into it, which is that you have to bring in content from your website, from your PIM platforms, from your ratings and reviews and wherever they're stored, right, in platforms like Bazaarvoice and so on and so forth. So that there's a lot of plumbing that had to come together and then there's the user experience, right? The consumer is asking a particular question, how do you answer it? What content do you present? Which products do you show? You're not going to show your 5,000-catalogue, you know, thing and, you know, how to personalize the experience for them. So that was the, you know, our vision was always, we didn't like the idea that brands had to settle for just a bunch of text representing the brand. We also didn't like the idea that, you know, if you're a brand and the consumer asks a question about sneakers, you, your sneakers with everybody else's sneakers just gets listed out there. And so, so Egenic Brand Store, which is the product that we launched, essentially uses—

You're getting there. Can I just interrupt you just so for everybody following along, if you use ChatGPT, so this is available right now with OpenAI, and you're looking at your screen where you typically type in your prompt, what you would do is go to more and you're going to see this new feature called apps that will pop up. And that apps is what Diaz is talking about now. So you're already going to see inside of there, Expedia is already there. Target is already there. Walmart is there. Adobe is there. So you're going to already see applications in there. And so when I go to chat and let's say, let's use Expedia. If I type in, I want to do a vacation to Hawaii. Ooh, I need a vacation to Hawaii right now, actually. And I mentioned that then Expedia could be triggered and bring in not just text but pictures as well. Right? Did I summarize that correctly? Yes, ma'am. Just so everybody's up to speed with us. Okay.

Exactly.

And now what you did is you have the ability to help brands showcase that through an app.

Right, exactly. Okay. So in a conversation, so you, you might ask a general question, like in your case, you're like, hey, I'm thinking of going on a vacation, maybe a quick getaway. I've got 3 days, you know, any ideas, right? And, you know, ChatGPT may say, hey, here's some destination to go to. You're like, okay, so what are, what are some of the hotels available in, you in, I don't know, Hawaii. And then it might say, well, there's these— there's a Hyatt, blah blah blah, there's a Marriott, whatever. That— and when you say, oh, I want to know more about that Marriott hotel or whatever, that's when the Marriott app will get triggered, right? And then now you have an opportunity to ask more specific questions. So that's the experience. And so what we're doing is basically generating that app for brands. So like Marriott and other brands so that from that point on, the consumer is conversing with the brand. It's highly curated, highly personalized, highly visual, you know, kind of like if you'd walk into a store for, you know, Armani or Prada or like, you know, it's that kind of experience, like a very, very immersive rich experience, but one that's driven by conversation. So it's not like here's 5,000 products, which one do you want? It's more like based on the question you asked, and, you know, and actually the other thing is context, right? Like, uh, these platforms remember, you know, um, what, what you tend to do. Like, do you tend to book luxury hotels, etc., etc.? So some context in there. So the answer is always a lot more intelligent, meaning the agents are doing the heavy lifting and presenting you with the right options at the right time so you can quickly arrive at a decision and book your hotel and go on your vacation.

It's really cool. So if you're a marketer, you know, what you're already seeing is that ChatGPT is really kind of building that funnel for you because someone's going to type in what they're looking for, right? Like, I want to buy this particular— I'm throwing a party, I need vodka, or I'm going to go to Hawaii, I need a hotel. So they're already almost pre-qualifying for you. So the funnel's getting smaller. And so now the capability that you guys have really innovated on, and you're the only ones I've heard of doing this so far for brands, is the ability to make that a visual experience so that you can see it. And if you're sitting there listening and you're like, I don't sell like a physical good, go check out the list. Like GoDaddy's there, Adobe's there. They're selling digital assets and they're already there as well. So don't be left behind. Just because you're like, oh, it's not me, right? So I know when you're looking at this today, part of what you guys do to build an AI-native experience, but that still uses the brand, you pull from PIM systems and libraries and PDPs and reviews, which maybe not all marketers know what those are. So what does that tell though a market about what type of content they now need to build, what type of structure, how they need to tag the assets even before they think about doing agentic commerce?

Yeah, that's a great question. The good news is most brands, especially commerce brands or brands that sell products to consumers, have a lot of that infrastructure in place already. So they typically have a— they have DAMs and they have PIMs. So the PIMs are sort of the last mile, which then sends that product content to retailers, right? So they'd send it to Walmart, they'd send it to Amazon, and so on. So that, so that's kind of where a Geniq brand store taps in. Or if it's a retailer, we would just tap into their product feed, right? They already have a product feed. And, and the idea there is initially just create discovery and visibility, right? Set up a storefront, your products are there, people can ask questions and so on., but then they can do a lot more. Like, for example, we're working with one large, uh, global luxury brand, so they're big on lifestyle, right? People tend to buy their products not because, uh, you know, it— they want to be part of that lifestyle, right? Um, and, and so they have a lot of content related to what to wear to a graduation or to a wedding or to your, you know, etc. And so it's almost like, hey, I'm going to, uh graduation, can you give me some ideas? And so they have all that content on their website. So we're starting to pull some of that content to essentially make suggestions or recommendations of products. And usually you'll see an image of a wedding or whatever, and then it's like, oh, these are the products that kind of match the current trendy styles if you're going to a wedding. So there's a lot of potential, but, you know, our advice in general is don't try to do it all that at once. You don't have to wait till you get all that perfect. You know, I, I come from the world of data warehousing where I remember I'd be like, oh, I gotta get all my data right before I start doing analytics. I was like, no, just get some of it out there, get people to experience the data, do the analytics, and they want more and you build it. So similarly here, there's enough content that brands and retailers have that you can immediately activate. So you have a presence because if you're not there, you're not there, right? Consumers can't find you or engage with you. And then there's plenty of opportunity to build richer, more engaging pieces of content that, that you have. It's all there. It's just sort of scattered in different places and needs to be pulled together and made available to these consumers.

So now all you marketeers who've, you know, worked so hard on your brand image, that can now come into Agentic AI. And in fact, it can even be more powerful because this conversational commerce really gets that intent data that someone's searching on that social really never could not get, right? They're kind of telling literally AI what they want and what they're confused about and what's stopping them from buying. So, you know, in your experience working with this, what should marketers be doing with that signal and should it be changing their campaign strategy?

Yeah, I think, uh, one of the things that's very interesting about how consumers engage or what they ask for is they often don't say, I'm looking for a Nike sneaker, or I'm looking for a pair of boots by brand. They often ask the question more in terms of what they are from a daily lifestyle standpoint trying to do. I'm going camping, 'What's a good hiking boot? It looks like it's going to rain this weekend,' right? That kind of question. Now you say, 'Well, they didn't actually ask for a brand.' And so how does your hiking boot associate with wet weather? Okay, and here's the interesting thing. Most product content doesn't say anything about that. It just says, 'Here's my image of my product, here's the price, here's a short description.' It doesn't say anything about wet weather. So that's the task ahead in a way, right? You have to think in terms of the questions consumers are asking about the purpose for which they may be buying the product— attending a wedding, going hiking, whatever. So there's some enhancement of the product content that needs to happen, and that can all be done. And the good news is AI can be used to do that, right? The old way used to be you needed to hire a whole bunch of people to sit there and look at each product— oh, is this a good product for rainy weather or not. So we're doing some really cool stuff there saying, hey, you know, we can scour, you know, for example, reviews and so on, and basically use AI to figure out that these particular boots seem to work well for rainy weather. Actually, it's very similar to what ChatGPT is doing too, if you think about it. They scour Reddit and YouTube and so on, and they try to figure out what are consumers actually talking about relative to a particular product or brand, and hence their answers are coming out. Now, that's potentially problematic for a brand because somebody put something on Reddit saying, "Oh my God, this was terrible. I bought it to go hiking in it," you know. So there's a little bit of that peril of leaving it to ChatGPT to represent your brand. But I think these agentic brand stores allow the brand to offer up those helpful tips saying these boots are great if it's wet weather, So I think that's something to think about. So there's the basics you can get up and running quickly. But the— from a how do we curate and manage that experience, we have to think beyond just product listings.

Yeah, it was really interesting because when I was chatting with Walmart in kind of this Q&A session, they're looking at, okay, I just had a baby, right? So I might go to ChatGPT and go, I just had a baby. This is my first baby. And so Walmart says, okay, well, if you've just had a baby, what's the first thing you're going to buy? Diapers, right? And so they, they start feeding that. And what they're looking at doing is saying, okay, you're going to go through a set of phases. So we want to remember that. So you bought diapers at this age, and then you're going to need, you know, medication that works for, you know, maybe a 3-month-old to a 6-month-old. I thought it was really smart the way they're thinking about those categories, like you said, boots with weather, how they're thinking about how are people talking in a conversational fashion about what you're doing. I use tools today, like I use SEMrush, one of our sponsors, to do that, to help me figure out what are the questions people are asking, because that then also can impact the content that you drive as well, right? Are you finding any other ways to figure out, like, what are consumers actually asking ChatGPT so that when that flag comes up that you're there? Like the weather question you just said, is that part of what you help brands do?

Yeah, exactly. So one of the cool things about the Genic Brand Store, once it's deployed, is that we get a lot of contextual data about how the the application got triggered based on a question that the consumer asked, right? So we're basically taking all of that data and turning it into analytics that can then advise the brand on, hey, here's, you know, here's how they— the app got invoked. It's because somebody was asking about the weather, or they were asking about— they just had a baby, or things like that. And that can help optimize the content that gets served up, like which types of content are more important, uh, and make these apps smarter at recommending the right products. So we do get that, those insights and the journeys, like do they first go and look at drink recipes and then buy the cocktail, or do they already know what they're going to buy and then want recipes for that? That's an interesting question, right?

It is, yeah.

But then we have the data for that, so as more and more consumers start using these apps, you can then say, oh wait, 60% of people actually are first trying to make a drink and then they're trying to figure out which cocktail, which liquor to use for that. So that's a big part of what this platform is doing also, is it's not just here's a bunch of content, here's how you engage with it. It's also helping the brand understand how to continue to optimize the content so that it becomes more and more likely the consumer's getting what they want. Quickly without having to go through a series of steps to get there.

I would say this is one reason why I always advise CMOs to have a Reddit strategy. At my company, we started— we have someone who— we have our own Reddit channel, but we also monitor other Reddit channels. Why? Because that's where you're seeing a lot of the questions that Chad and Claude and others are pulling in. So it really helps you to have a digestible, I think, strategy for helping figure out, like, what's going to happen there. So, oh, I'm sorry, go ahead. Did you want to say something?

No, I was going to say, absolutely right. I think there's also, you know, retailers and others are implementing chat on their own sites. And so some of what's being discovered from there can and should be fed back into the LLM environments also. I think for most brands and retailers, the biggest challenge right now is sort of controlling the the conversation content, because it's great that ChatGPT gets it from Reddit and YouTube, but if I were a brand, I'd be worried about being represented by ChatGPT versus me being able to represent my brand, the consumer, right? So, so I think the more of that data and experience they have and they can publish it in, the better it is for the brand.

Now this is so exciting. I was actually just with Accenture this morning and Accenture has now put investment in Agentic Commerce and in fact in your company. And I think that makes all enterprise marketeers pay attention. What do you think that message that sends to a CMO who's running pilots and waiting to see how this plays out now that Accenture has jumped in? What do you think the big signal there is?

Yeah, you know, it's interesting because Accenture has a very, uh, probably one of the best vantage points from which to look at the market, right? They work with all the global large brands and retailers, right? And so I think their interest in agentic commerce represents what they're hearing from their clients, right? And it's an aggregate of, you know, 9,000 clients that they work with. So talk about market research, right? They have the best market research. Yeah, what could have— one of the questions I always get when I get on a call with a brand or a retailer is like, is anybody else doing this? It's like, no, we want to be first. Well, some of them ask like, no, I want to be first, like how quickly can I get this up and running? So we love that. But others be like, hey, so I think Accenture saw this probably ahead of almost everyone else because they were having conversations with clients and the clients like, oh my God, there's all this stuff happening and Labs, we feel like we're not there. What do we got to do? So that's what I think is driving Accenture's interest and the investment they made in DaVinci Commerce, because they, they do believe in this agentic commerce model and are hearing it from clients. And, and therefore, I, I feel like, you know, there's a lot in the media about what's hot and what's not. You know, I always say, hey, you know, when customers say, express their worries or what they see as the next way to engage with consumers, then it becomes real. I think for us, the Accenture partnership is fantastic because they have a deep bench of skills in each of the verticals for different brands and so on. You know, we're an amazing AI tech platform company. We know how to build AI, you know, platforms really well, but I think it's that combination of their expertise, understanding client requirements, and our technology that delivers great solutions to these brands.

I think that's very well said. So Diaz, I know we're running out of time here, but my big question is, you know, Open Claw has come on the market and now, you know, I have my own Open Claw running over here and it's buying things for me. So how does this change, like when agents start purchasing on your behalf, things change and what they want to see. How does what we're talking about today with apps, how does that change with an agent?

Yeah, that's, that's quite, uh, interesting. And I think, you know, the idea that an agent could go buy something from you is something that, you know, uh, depending on who you ask, either think it's going to be the future in it already is there, or, oh my God, I would never have an agent do that for me. I think the answer is really, it depends on the type of product, right? We, we all buy certain products over and over again. I wish I had an agent. I could just tell an agent to do my grocery shopping every week. It's just like, I buy this stuff, I don't really need to sit there and click and click and, you know, do all that. So I could definitely see some great potential there. I could also see agents play a role in, uh, I was surprised actually in talking to some, uh, brands. They, they start talking about customer service. It's like, wait, we're talking about e-commerce here. Like, no, actually somebody wants to know where their order is right now. Uh, is it on its way? Where is it? You know, when is it coming and stuff? So these agents play an important role also because it's not just about surfacing the content, because when you bought something, you're like, when is it coming? So I think these agents can actually do stuff for the consumer, right? Can you go check on my order? I want to know where it is, what's going on. So agent goes, checks the FedEx, you know, routing number, and comes back and says it's on its way, it's going to arrive this afternoon, whatever. So I think the potential for Open Claw-like eGenic technologies— I know Claude announced something this morning where there's lots coming out. Yeah, yeah, that's sort of the, the next level. Once you've got curated content in front of consumers and you figured out how to get them buying stuff. There's several other applications that come into play where, you know, as a consumer, I can say I'm kind of lazy, I don't want to do a lot of work, so I'd rather just go do it for me, but do it intelligently, right? Don't go buy stuff I would never buy, or, or use my credit card and go shopping based on what you think I need. But there's a lot of intelligence these agents are starting to have. So it'll be interesting to see where it goes, but I don't think you can outright dismiss the idea of agents doing it for you. It may take some time for consumers to get comfortable with it.

Yeah.

Much like even shopping on LLMs, right? I think the early studies were like, oh, they don't trust the LLMs to do it for them. I think we'll get there and it probably won't be too long judging from the speed at which OpenClaw and other agentic technologies are being adopted.

Yeah, it was fascinating because when I was talking to Walmart again, they gave me an analogy that I thought was really good about the difference between an agent and a person. They said, you know, for a person, you might price things at $9.99 because there's a barrier, right? $9.99 versus $10 to a human has an emotional impact. But to an agent, it's one cent. And it really doesn't have the same emotional boundaries. And so I thought that was quite fascinating, just kind of looking at that, that little difference. I was like, oh, that's true, right? Like, I might buy something for $9.99, but not $10. Um, maybe silly, but true. So I do think there's going to be some more work done there too. So, um, to close this out, if a marketer is listening right now and they wanna stop being a bystander in this agentic commerce area, what is one thing you think they should go do this week? You know, even before they have a budget or they select someone like you to help them, what should they go and do?

Yeah, I would say, you know, create a way to experiment and learn, right? Everything's changing so fast and you know, what's true today may not be true next week or next month or whatever. Instead of standing on the sidelines where you're not learning much about what's actually going on, I think making an investment to say, okay, I'm gonna make a bet, even if it's a small bet, it's the learning that's more valuable than even the bet you made. If you stand on the sidelines, by the time everything settles and everything is known, you as a brand are going to be way, way behind because it's a lot of work to understand behaviors, to prepare your content, to get it into these platforms, to build the agents, to do all the heavy lifting. So the sooner you understand it and the quicker you get in the game, the more likely you are to at least keep pace with the market. And for the ones who jump in early, they will likely be ahead of, of their competitors and peers. So I would say just do something, even at small scale. At the very least, you'll learn a lot about what you need to be doing at bigger scale.

I think that's great advice. At Amazon Web Services, we would say that that is being curious, right? Being curious, exploring, and learning, which is a skill set that will never go out. Diaz, thank you. I will sum up, I think, 4 things that everybody on this call needs to know. One, the storefront has moved. Whether you're selling a digital asset or a physical asset, consumers are now starting to shop, and they're starting that shopping journey inside the LLMs. 2, brand control is now a technical problem. So you can't style guide your way into an LLM. So that content architecture that Diaz talked about is super important. We ended with your next customer might be an agent, might be Open Claw. These autonomous agents will shop and purchase on your behalf. What do you need to change about that? And then I think the grand finale that Diaz walked us through was conversational intent and that data changes a lot of stuff that you have to do. So that learning and being curious will play a huge role in how you present this in the future. Thank you, Diaz. Thanks for sharing the announcement from DaVinci. And I really appreciate you as like a phenomenal CEO having multiple exits, sharing your insight to marketeers. It's always great for marketers to hear directly from a CEO. So thank you so much for that.

Yeah, thank you, Sandy. Always a pleasure to chat with you.

Great to chat with you. And as always, make sure you do a big thank you to SEMrush because they're doing a lot of the stuff today. Check them out as well as Brevo. I'll have all of that in the show notes. Thank you so much. Join us next time.