Good evening.
I’m not sure how much credit I really want to give him because I don’t want it going to his head, but I will admit that the serial entrepreneur Scot Wingo did play at least some role in my interest in starting The Aisle. For those who don’t know him, Wingo is a longtime e-commerce entrepreneur probably best known for co-founding and running ChannelAdvisor, the e-commerce software firm that went public in 2013 and that helped brands sell on Amazon, eBay and other online marketplaces.
He’s also the co-host of the smart industry podcast, The Jason & Scot Show (I’m not that Jason, though I have appeared on the show a few times, most recently when I launched The Aisle). More recently, he started Retailgentic, a podcast focused on agentic commerce.
About a year ago, I checked in with Wingo to hear about what he was building with his new startup, ReFiBuy, and his excitement for the AI era of commerce was contagious. I had by then caught a bit of the bug myself, but I found it reassuring that a grizzled industry vet was maybe even more excited than I was about the opportunities, challenges, and overall impact that this new era of AI would have on online shopping and retail.
But Wingo’s excitement told me something else, too. If a serious, respected voice like his was this fired up about the AI era of commerce, there would inevitably be others looking at the same moment in more opportunistic, even exploitative, ways. As a result, there would be a need for rigorous, independent reporting on a space that was about to get very noisy. Some signal amid the hype. And here we are.
This seems especially relevant now because Wingo’s startup, ReFiBuy, announced a $13 million+ seed investment this week. Wingo’s pitch for the startup is that its technology helps brands refine the product information that AI search engines rely on for recommendations, send that information to AI platforms and the brand’s own site, and then check how those products actually show up in AI-generated shopping results going forward.
Wingo’s view is that AI shopping makes almost every product search unique—”there is no head/tail anymore…it's all very long tail conversations where your products are surfaced”—because each shopper asks questions in their own way. As a result, the bet is that brands will need a system that’s continuously updating product information based on what AI tools are actually showing shoppers.
And, yes, ReFiBuy is practicing what it’s preaching.
“We're replicating what a human would do and we had hundreds of humans doing at ChannelAdvisor back in the day,” Wingo told me, “but now with an army of 24x7 agents.”
Also on my radar right now:
Anthropic expanded its version of apps for Claude, called Connectors. I’ve been skeptical of the future of similar experiences on ChatGPT, but the ability for a Claude user to invoke a relevant consumer app experience (say Instacart’s or Uber’s) just by asking a relevant query has me paying attention.
Speaking of OpenAI, ChatGPT officially launched its self-serve advertising suite for marketers today. Marketers, keep those first impressions incoming.
Finally, one of the most stunning stories in e-commerce right now is that eBay is back to real, true growth mode. Bad timing for GameStop’s ridiculous acquisition bid.
“Could you imagine if someone tried to tell us a few years ago eBay would be growing at this rate?” one subscriber marveled in an email to me. “I would have checked them into an institution.”
Now on to the good stuff…
The Center Aisle
For anyone tracking where AI might disrupt commerce most aggressively first, grocery shopping seems like one of the categories most ripe for change—at least when it comes to making re-ordering staples or favorites easier, or autonomously comparing prices across stores (as these startups I wrote about are trying to do). There are also more straightforward chatbot use cases, like prompting for recipe or diet ideas.
Instacart has been experimenting aggressively in this space, and has built dedicated experiences that can be accessed inside ChatGPT and Claude. DoorDash has been talking up AI-powered discovery as well. Kroger, Albertsons and Walmart are working on their own flavors of smarter grocery shopping, and Amazon isn’t sitting idle either.
Then there’s Hungryroot, the 11-year-old online grocery company that reportedly did $700 million in revenue last year and has spent years using older-school AI to help recommend groceries, plan meals and fill customers’ digital carts.
The company’s core system relies on classic machine-learning models, structured data and business rules to score products, rank recommendations, manage budgets and account for inventory.
That may sound less exciting than a chatbot that promises to plan your meals from a prompt. But CEO Ben McKean’s bet is that it will matter more.
Subscribe to The Aisle to read the rest.
Become a paying subscriber of The Aisle to get access to this post and all my work. Plus some perks.
