Latest / Retail Media Breakfast Club / The Coming AI Marketplace Gold Rush
Transcript
- The Coming AI Marketplace Gold Rush === [00:00:00] I remember the ads on YouTube and you probably do as well start an Amazon FBA business. You'll make millions of dollars and you know, they weren't entirely wrong [00:00:15] between say 2012 and 2020. Thousands, if not tens of thousands of entrepreneurs built businesses entirely within Amazon's walls. No website, no [00:00:30] retail presence, no marketing infrastructure, just an Amazon seller account, a product that they found off of Alibaba and a dream. [00:00:41] I watched this ecosystem explode firsthand. [00:00:45] An FBA business became a noun, a business model, a path to wealth for the middle class. Some brands eventually transcended the platform like Onca [00:01:00] zesty. Pause. Kaori Hero Cosmetics. It's actually a pretty short list, by the way, but some of those companies did eventually transcend the bounds of Amazon and secured [00:01:15] nationwide distribution. [00:01:16] But for every breakout success like that, tens of thousands of sellers remained wholly dependent on Amazon for everything, not just for the warehousing and fulfillment, but the customer [00:01:30] acquisition, the marketing. Processes and even their business identity. Over time, the music faded. Amazon opened the floodgates to Chinese manufacturers advertising costs sawed, [00:01:45] organic visibility vanished, and those YouTube courses disappeared and those YouTube ads disappeared as quickly as they'd arrived. [00:01:56] The Amazon FBA gold rush was [00:02:00] over. Now we could see the same pattern emerge with AI enabled checkouts. Except this time the cycle will be even more compressed. Let's jump in [00:02:11] [00:02:13] the Amazon FBA [00:02:15] Boom was beautiful in its simplicity. [00:02:17] entrepreneurs could outsource their entire operation to one platform. Amazon stored your inventory, shipped your products, handled returns, provided customer service, and most [00:02:30] crucially delivered customers. You didn't need a warehouse or even a website, just products, and your Amazon seller account. The early movers printed money, although probably not as much as the ones selling the [00:02:45] courses that stoked the industry. [00:02:47] With relatively few sellers and Amazon actively promoting this third party marketplace, organic visibility was achievable. A decent product with good reviews could eventually be [00:03:00] discovered without massive ad spend, but marketplace dynamics are predictable. As supply flooded in discovery became pay to play, amazon realized they could build more supply by welcoming in direct relationships [00:03:15] with factories in Shenzhen, rather than continuing to appease suburban middlemen operating businesses out of their basements. And as this supply flooded in discovery became pay to play [00:03:30] as well. [00:03:30] Amazon discovered it was sitting on a gold mine with its nascent ad business, and by 2018, Amazon was pocketing 40% of sellers revenue. According to Marketplace Pulse. By [00:03:45] 2022, it was 50% The arbitrage window had closed. So here's the history lesson. Marketplace booms are timing. Arbitrages profit exists in that sweet spot between when a platform opens up [00:04:00] and when discovery becomes expensive. [00:04:02] So now let's think about what this AI marketplace would look like. Right now, the infrastructure for AI powered commerce is taking shape. Although Open AI launched [00:04:15] chat GBT five last week and didn't include a branded checkout as many pundits had expected, I still don't think we're too far off from that reality. [00:04:26] Kelly Girt, COO of Tech [00:04:30] Company. Pipe 17 said on LinkedIn that when OpenAI releases its own branded checkout, it will take a transaction fee from merchants just like any other checkout does today. Think about that. Chat. GBT [00:04:45] won't just recommend products to you. [00:04:48] It will complete the transaction without you ever leaving the chat. The LLM itself becomes the storefront, [00:04:56] And I've talked about this before in terms of agents going [00:05:00] to a retailer and checking out or hooking into MCP technology to make a checkout, but there is a scenario where the LLM itself becomes the storefront and the checkout and the payment processor [00:05:15] and the marketplace all sits just in one layer without external calls to retailers at all. [00:05:23] But while it took many years for the general population to get used to the idea of entering their credit [00:05:30] card details, to buy stuff on the internet, AI commerce will be able to draft off all of those existing e-commerce behaviors and. Widespread adoption of ai. [00:05:45] Acosta group's, retailer intimacy is legendary. Merchants answer their calls. You just can't find that access at the same scale with any other partner [00:06:00] out there. [00:06:00] That expertise with both retailers and Shoppers fuels its Connected Commerce team, which offers digital shelf retail, media, and data analytics all under one roof. Tap into a hundred years of [00:06:15] retailer relationships and award-winning digital commerce capabilities. Learn more at ACOA Group and tell them Retail Media Breakfast Club sent you. [00:06:25] That's A-C-O-S-T [00:06:30] group. So in terms of a timeline, I really liked what Carl Haller, who is a partner at IBM Consulting, had to say around this timeline on an episode of the [00:06:45] Retail Genetic podcast. [00:06:46] He spelled out this timeline, which is that this year, 2025. Really it's just the early adopters who are going to use shopping agents. For holiday shopping next year, next [00:07:00] holidays, especially 2026, that's when there's gonna be more mainstream usage of shopping agents. By 2027, agentic shopping will become the default e-commerce behavior. [00:07:14] [00:07:15] Carl made another interesting prediction on this podcast. He said that AI native brands are coming, and ~he says that AI native brands are coming.~ [00:07:23] ~He says ~that AI native brands could be much larger and more profitable than traditional [00:07:30] companies today, and that's because they could build every function around autonomous agents from day one. Everything from product design, merchandising, supply chain finance, customer service, all could be orchestrated by AI with [00:07:45] minimal human oversight. [00:07:46] And he lays out a case to say that a fashion business would traditionally need 75 to a hundred people to reach a hundred to 200 million in revenue. But [00:08:00] with an AI enabled fashion business, you could achieve that same scale. With just 20 to 30 people, and when your gross margins stay fat because you're not paying for that headcount, you can [00:08:15] reinvest all that cash flow into growth. [00:08:18] Just as FBA brands road, Amazon's infrastructure to scale. AI native brands will ride the agent infrastructure, but the scale is different. Where FBA enabled entrepreneurs to [00:08:30] build 10 to $50 million businesses with small teams, AI marketplaces could enable. [00:08:37] A hundred million dollar plus businesses with similar headcounts. Now, it's not all [00:08:45] sunshine and rainbows, of course, when shopping happens through AI agents, brands could lose even more control of the customer transaction than they did with Amazon,. at least with FBA sellers could optimize [00:09:00] listings, manage reviews, build some kind of brand equity within Amazon's walls. But with AI agents, the agent is the one that decides what to show the customer. The agent owns the relationship. The agent controls discovery, [00:09:15] and that's why Jason Goldberg, chief Commerce Officer at Publicis. [00:09:20] Sums up the risk to incumbent brands as follows, he says Retail brands that used to build moats with the strength of their brand and their [00:09:30] direct relationships with shoppers get reduced to commodity suppliers, all competing to win reverse auctions for the lowest margin sale possible. The very efficiency that would make AI native brands possible could [00:09:45] also make them fleeting. [00:09:46] Why this window will slam shut even faster. The AI marketplace opportunity will compress from years to months. For one simple reason. Everyone learned from Amazon FB. A sellers [00:10:00] enjoyed years of organic discovery before pay to play kicked in. AI platforms won't make that mistake. We should expect algorithmic bidding and embedded ads from day one. [00:10:12] When Carl Haller talks about a 20 [00:10:15] person team reaching a hundred million dollars in revenue, he's describing a world where those same efficiencies are available to everyone else instantly. No learning curve, no infrastructure. So this arbitrage window that took five years to close. [00:10:30] With the Amazon marketplace, opportunity should slam. [00:10:34] Much faster with AI marketplaces. The YouTube ads aren't here yet, but I can just picture them in my head. Build an AI native brand. [00:10:45] Let chat JBT Sell your products. The future of commerce is conversational. And just like FBA, there will be real winners, brands that use AI marketplaces as a launchpad, not a permanent home.[00:11:00] [00:11:00] Companies that build genuine differentiation and become cult favorites. The UNCs and hero cosmetics of the A IH, but for most the same fate awaits, the window will close, margins [00:11:15] will compress, and then the YouTube gurus will disappear as well. But that gold rush is likely to last only months, not years.