Amazon's agentic army is coming - you ready?

This, this is the Billion Dollar Sellers Podcast. Your go-to source for cutting-edge strategies and success stories from the world of Amazon and e-commerce. Buckle up and get ready to take your Amazon business to new heights. Don't forget to subscribe to the Billion Dollar Sellers newsletter. Welcome your host, welcome your host, Kevin King. >> Hey everyone, welcome to the Billion Dollar Sellers Podcast. I'm your host Kevin King and today is May 11th, 2026. So, we got a big show for you today and honestly, this one's going to get you thinking about the future of how people actually buy stuff on Amazon. We're going to talk about Amazon quietly building a whole department, 40 engineers, to plug their commerce into outside AI apps like ChatGPT and Claude and Gemini. Then we'll get into Rufus and how Amazon's testing blending AI answers right into the regular search results, which is a big deal for sellers. Also, Walmart just fired OpenAI's instant checkout after only 5 months and replaced it with their own AI assistant called Sparky. Plus, we've got some wild stats, uh, software tool of the day, subscribe and save changes and more. All right, but first, here's your stump Bezos question for today. So, the average US-based Amazon seller does about $375,000 in sales per year on Amazon. The question is, how many sellers do over 1 million per year on Amazon? Think about that and I'll give you the answer at the end of the show. All right, let's get into it. And so, uh, Amazon is quietly building what I call an agentic army and this is a big For months now, Andy Jassy has been saying that Amazon plans to partner with outside AI companies and plug parts of its shopping experience into their products. And, you know, talk is cheap, right? But, hiring isn't. Last week, Amazon posted a job listing for a principal technical program manager to run, and I quote, "a dynamic engineering team responsible for Amazon's strategic integration with third-party agentic platforms." And this role oversees 40 engineers. That's not a pilot, that's a full-blown department. So, a guy named Joe Kaziukonis spotted the listing first, and then Jason Del Rey over at the AI connected it to a bunch of other moves that really paint a clear picture of where Amazon's going with all this. So, here's the thing. Until recently, Amazon's AI shopping playbook looked pretty defensive. They launched Rufus, blocked AI crawlers from the site, and even sued Perplexity for scraping. Very defensive, you know, classic Amazon flywheel protection. But then last week, Amazon joined the Tech Council for something called the Universal Commerce Protocol, which is a new standard designed to let AI agents shop across retail and AI apps without needing custom integrations for every store. And now this job listing, both point in the same direction. Amazon's preparing for a world where shopping starts inside AI products that Amazon doesn't own. So, the job description hints at two possible plays, and Amazon might run both. The first one is the distribution play. Yeah, so think about Amazon's whole stack, the catalog, Prime shipping, reviews, checkout, all of that living inside ChatGPT, Claude, Gemini, or whatever comes after that. You ask an AI to buy something, and Amazon fulfills it behind the scenes. This one feels like a question of when and how, not if. The second is the merchant play. Amazon recruits Shopify stores, DTC brands, other retailers into its buy for me agent, turning what's currently a scrappy experiment into a real channel for buying from outside Amazon's marketplace. The listing doesn't say which comes first, but the scope of the role says both are on the table. So, why should sellers care about this? If Amazon succeeds at injecting its commerce layer into outside AI apps, the real estate that matters shifts. Your product page on Amazon still matters, but the surface where the buying decision actually happens might be a Claude conversation, or a ChatGPT recommendation, or an agent acting on somebody's behalf inside an app that has nothing to do with Amazon. That changes what ranking means. It changes what reviews do. It changes whether your A+ content even gets seen before somebody hits buy. And it raises a question Amazon hasn't answered yet. When AI agent on somebody else's platform buys your product through Amazon, who owns the customer data? Who gets credit for the conversion? Who controls the experience around the sale? Amazon clearly wants to be the one controlling where, how, and with whom its commerce shows up beyond amazon.com. And that 40-engineer team is being built to make sure of it. Watch this hire. Whoever takes that job is going to shape a meaningful chunk of how product discovery works for the next decade. If you want to read Jason Del Rey's full piece on this, there's a link in the show notes. All right, let's talk about some interesting stats. So, Marketplace Pulse put out their latest breakdown of Amazon's net sales by category, and a couple of things really jump out. Amazon Web Services has gone from about 13% of revenue back in Q4 2021 to almost 21% in Q1 2026. That is a massive shift. The cloud business is eating a bigger and bigger piece of the pie. Advertising is also been creeping up. It was about 7% a few years ago, and now it's pushing almost 10%. So, Amazon's ad business just keeps growing as a share of total revenue. On the flip side, online stores, which is Amazon's 1P retail, has dropped from about 48% down to about 35%, and 3P seller services is sitting around 23%, pretty steady. So, the takeaway is that Amazon is becoming less of a retailer and more of an infrastructure and advertising company. That's just, you know, that's where the money's going. And then Shopify just hit a pretty wild milestone. Q1 2026, their gross merchandise volume crossed $100 billion for the first time. That's a 35% jump year-over-year. Their total revenue hit $3.17 billion, so that's up 34%. And their operating income jumped 88% year-over-year to $382 million. So, Shopify is not just growing, they're getting more profitable as they scale. That's a pretty healthy sign for the DTC ecosystem. All right, now let's talk about Rufus and Amazon search cuz this one's really important for sellers. So, Amazon is signaling that they could start mixing AI-generated commentary directly into the main search results. They're blurring the line between Rufus and that traditional search bar. Amanda Door, Amazon's VP of core shopping, said they're testing hybrid modes where an AI blurb shows up above your search results for certain queries without pushing you into a separate chatbot. Some shoppers have already seen it. AI summaries with product recommendations with an option to keep chatting. And it comes down to the type of query. For simple stuff, the traditional grid wins. If you're searching for milk, you just want to see the milk you usually buy, how much it costs, add to cart. You don't need a paragraph about 2% versus whole milk. But for open-ended stuff like, let's say, researching hiking boots for the summer, that's where Rufus makes sense. And here's why Amazon is moving carefully on this. The search bar powers most of their online retail revenue and it underpins their $68.6 billion ad business, which is mostly sponsored search and product ads. You mess with that flow, there's real risk. Google, by contrast, has already gone all in with AI overviews and AI mode. Now, here's some Rufus numbers that are pretty wild. Amazon says Rufus generated nearly $12 billion in incremental annualized sales in 2025 since its launch. That's not nothing. And when shoppers use Rufus, they're almost three times more likely to buy, 2.74 times to be exact. The conversion rate goes from 21% with no Rufus to 35% with one to three interactions, 47% with four to 10, and 58% if they go 11 plus. Holy cow. A few other things worth knowing. Grocery is getting its own playbook. Doug Harrington, the retail chief, has made perishables a priority and Door's team is building a grocery experience that looks different from the rest of the site. Search for milk and Amazon wants to surface cereal and bananas to help you build a basket rather than showing you 25 different milks. Just tap tap tap add to cart. That's the goal. Auto completes is doing heavy lifting, too. Nearly 60% of US searches now come from a shopper accepting an auto complete suggestion. And Amazon recently added thumbnail images next to those suggestions. Visual search is expanding. They got this thing called Lens Live. It lets you scan your surroundings in the app without taking a photo with Rufus integrated for follow-up questions. And outside agents are still locked out. Andy Jassy reiterating in January that outside shopping agents aren't welcome on Amazon. They're citing poor customer experience and unresolved economics. So, what's the takeaway for sellers? Amazon isn't going to replace search with chat, but they're going to keep stitching AI into the existing flow wherever earns its place. Your listings still need to win the grid for those transactional queries, but the conversational layer above search is coming. And it's going to reward products that have clean structured data that AI summaries can pull from easily. All right, time for the BDS and software tool of the day. So, today's tool is charm.io. And this one is really interesting if you work anywhere in the DTC or TikTok shop space. Charm is a data platform that tracks daily sales estimates, growth scores, and verified contacts across over a million online stores and TikTok shops. Think of it as a searchable index of the entire DTC and TikTok shop universe. They're tracking over 4 million direct-to-consumer brands across social engagement, web traffic, product pricing, ad exposure, revenue growth, all of it. Their TikTok shop coverage spans the US, UK, France, Italy, Germany, Spain, Mexico, Brazil, and all of Southeast Asia. The daily sales data on shops, products, creators, hashtags, and videos. The core feature is their prospector tool. You filter brands across over 160 metrics, things like growth scores, ad counts, tech stack, marketplace presence, social media metrics, and you build targeted lists. From there, you can pull verified contacts, push data into a CRM, or hit their API. Agencies use it to hunt DTC clients, retailers use it to scout brands to stock, investors use it to find breakout brands before they pop. And 3PLs and SaaS companies use it for prospecting. If your work involves finding, ranking, or pitching DTC and TikTok Shop brands, then Charm is the data set most other tools are missing. You can find a link in the show notes. All right. This next one is uh I think it's one of the biggest stories this week. Walmart just killed its partnership with OpenAI's instant checkout after only 5 months. Remember, this was supposed to be a big deal. Doug McMillon and Sam Altman both hyped it up. Customers could shop Walmart inside ChatGPT, but it just couldn't deliver. Conversion rates fell well below what Walmart sees on its own channels. The feature struggled with product accuracy and couldn't sync with Walmart's internal shopping tools. And the deeper problem was how instant checkout was built. It scraped retail websites for product data. Scraping tells you what exists, but it doesn't tell you what's actually in stock, what shipping looks like, or what's accurate right now. Emily Pfeiffer from Forrester put it bluntly. She basically said that crawling and scraping isn't enough to run real commerce. And Bob Hetu from Gartner said that OpenAI underestimated how hard transactional enablement actually is. So, what will Walmart do? They pivoted. They are now embedding their own assistant called Sparky directly inside ChatGPT and Gemini. ChatGPT Plus and Gemini advanced users got access in March. Free tier is coming this spring. And reportedly, Anthropic is next, so Sparky might land inside Claude 2. And the early numbers explain the move. Sparky users inside ChatGPT are converting at roughly 70% of the rate of direct walmart.com shoppers, which sources say is far above what instant checkout was doing. The EVP of AI acceleration at Walmart, Daniel Danker, told investors that instant checkout was a very temporary moment in time and that within a month customers wouldn't even see it anymore. And Walmart isn't alone. Target, Instacart, Shopify, and Etsy are all building their own embedded experiences inside AI platforms instead of handing the keys to OpenAI. So, the agentic commerce land grab just shifted. The early bet was that ChatGPT becomes the storefront and retailers feed it data. But, the new bet is the opposite. Retailers control the agent and the AI platform just becomes a distribution channel. So, what does this mean for sellers? If you sell on Walmart Marketplace, then your product data, inventory accuracy, and listing quality now matter inside Sparky the same way they matter inside Walmart search. Sparky pulls from Walmart's verified catalog, not scraped snapshots. Clean data wins. If you sell on Amazon, watch this closely because Rufus is Amazon's version of the same play. It's the same logic, the same incentive to keep customers inside the walled garden. The bigger signal here is that scraping-based AI commerce just took the public loss. Verified first-party data won the first real test. The sellers who treat their listings like a database, not a billboard, going to keep getting picked. Quick break. If you're selling on Amazon and you're not using micro influencers to boost your rankings, you might want to check out Stack Influence. Stack Influence lets you automate micro influencer product seeding at scale. You send free products to influencers, they post about it, and you get increased Amazon rankings, UGC content, and more external traffic. Big brands like Magic Spoon, Unilever, and Mary Ruth Organics have used it to get to page one and increase monthly revenue by as much as 13 times in as little as 2 months. That's wild. You pay influencers only with products, no negotiating fees, and the whole thing is 100% automated. You can get 10% off if you sign up this month. There's a link in the show notes. All right, this one's for all you CPG sellers out there. Amazon just made a change to the Subscribe and Save dashboard that you need to know about. The team at Incremental Digital flagged this. Amazon added a side-by-side view that shows manual reorder revenue alongside subscription revenue. And why that matters is that subscriptions are recurring, they're predictable, and they produce longer customer lifetimes than manual buy again reorders. Before this, brands were basically blind to the manual reorder side. Now you see exactly how much money you're leaving on the table. Incrementum gave the example of a brand hitting a 32% reorder rate with low single-digit subscribe and save. On a $50,000 a week business, that's about $16,000 a week from loyal manual reorders. You convert even a third of those to subscribe and save, that's roughly $5,000 a week in recurring revenue with zero new customer acquisition. Holy cow. So, here's what you should do this week. First, sweeten your subscribe and save offer. A weak base discount with no coupon is the top reason loyal buyers don't subscribe. Use the dashboard to see what's converting in your category and test a higher tier. Second, build retargeting flow. Create an AMC audience of customers who reordered but never subscribed and retarget them with sponsored display. One important thing though, Amazon sponsored display creative policies don't allow specific dollar or percentage saves claims on custom imagery. So, lean on convenience messaging, things like never run out instead of discount language. Third, use a gallery image for your subscribe and save benefits. Secondary images are one of the few spots where Amazon lets you highlight subscribe and save eligibility and value. Swap a product shot for an infographic. And finally, prioritize the right ASINs. Find SKUs with high reorder rates but low subscribe and save adoption. That's where loyalty already exists and only the lock-in is missing. Start there. The new data doesn't change retention math, it just makes ignoring a lot harder. Okay, before we wrap up, a few more hot picks for you. Whatnot just integrated with Shopify, so if you're doing live selling, that's a big one. TikTok shop sales hit $4.9 billion in Q1. It's a big number. And $300 billion dollars US imports which country of origin in 2025. So, if you're doing any importing, uh that's something you want to keep an eye on. Links to all of those in the show notes. And here's your parting shot for today. You can't control how you feel, but you can always choose how you act. That's from Mel Robbins. I think that's just a good reminder for anybody running a business. You're going to have bad days, stuff's going to go wrong, but it's what you do about it that matters. Oh, and about that stop Bezos question from the beginning. How many Amazon sellers do over 1 million dollars per year? The answer is 11,000. 11,000 sellers doing over a million a year on Amazon. That's a lot of people crushing it. All right, that's all for today, folks. Um I'll see you again on Thursday. This is Kevin Tayne signing off from the Billion Dollar Sellers Podcast.