Stitch Fix was an OG in AI. Can its comeback stick?

Good evening.

Before I get to today's main column, I admit that I’m still trying to digest a very weird week at ChatGPT parent company OpenAI…and what, if anything, it means for all of you.

The quick rundown for those not living and breathing this every day: On Thursday, April 2, the company announced the acquisition of TBPN, a young, niche, but popular-in-Silicon-Valley media company that runs a three-hour tech talk show streaming live on X and YouTube every weekday. The acquisition price was reported in the low hundreds of millions. A lot of digital ink has been spilled over OpenAI's strategy here, but my initial gut feeling about the deal—besides checking if it was still April Fools’ Day when it was announced—was that it sounded like what a perhaps-unfocused, flush company would do if it wanted to purchase something "cool" basically because it could.

Then on Friday, OpenAI announced its CMO and its "CEO of AGI deployment" were both stepping aside for health reasons—the latter temporarily. On Sunday, The Information reported that CEO Sam Altman and CFO Sarah Friar might not be aligned on the company's readiness to go public. And on Monday, The New Yorker published a deep dive on Altman that raised questions about his trustworthiness, but without surfacing any smoking guns that put his job at immediate risk.

Phew.

So how does all of this affect whether and how retailers and consumer brands should think about partnering with OpenAI? My gut: not dramatically, but the executive turbulence is hard to completely dismiss. If credible reporting is suggesting the CEO and CFO aren't on the same page—publicly, they are saying they are—and other key leaders are stepping aside, you have to wonder how much bandwidth is actually being dedicated to retail and brand partnerships down the chain.

For now, I’d echo the advice of several top retail tech execs I spoke with at Shoptalk—including leaders at The Home Depot and Lowe's—who said they'll keep experimenting with AI platform partnerships, but without diverting resources from near-term, clear-win initiatives. Right now, that cautious posture looks pretty smart.

***

I’ve also been watching early signs of brewing merchant unrest at Amazon, following three rapid-fire hits to third-party Amazon sellers, who account for about 60% of units sold in the Everything Store: a new 3.5% fuel surcharge for many North American sellers; a shift to Amazon holding onto seller payouts until seven days after a customer delivery; and new restrictions preventing many sellers from using credit cards as their primary method for buying ads on Amazon.

All of these add up to a cash squeeze on these SMBs and other sellers I’ve heard from. Amazon had to know that such a combination of changes would cause some outrage, but its communications with its seller base have been notoriously poor for so long that it's hard not to wonder if that's by design. The alternative—utter incompetence—is tougher to believe for a company as successful as Amazon is.

So what does this have to do with AI? I’m glad you asked. It’s hard to watch Amazon squeeze its least powerful partners on cash flow without thinking about the enormous infrastructure costs the company is taking on to fund its AI and cloud ambitions—both for itself and for enterprise customers. CEO Andy Jassy has said that number could reach $200 billion—with a B—in 2026 alone. Whether there’s a direct link between those costs and the current seller squeeze or not, it’s difficult to ignore that backdrop.

Speaking of AI, I had to chuckle when I saw the responses from Amazon’s own shopping assistant, Rufus, to questions concerning the triple whammy of seller changes.

Now on to the good stuff…

Some of my earliest, fondest memories of covering the e-commerce industry involve chasing stories about Stitch Fix, the personal styling service founded in 2011 and led for its first decade by the entrepreneur Katrina Lake.

In 2014, I remember hounding legendary venture capitalist Bill Gurley during the Code Conference at the Terranea Resort, trying to get him to confirm a new funding round for the then-young startup. He didn't bite, but I broke news of the $30 million investment eventually. Two years later, I cornered Jeff Bezos backstage at another Code Conference and asked about a new apparel business Amazon was entering with its own array of private-label brands.

After I seemed to unintentionally insult the Amazon CEO by asking him if he was wearing any Amazon-branded clothing himself, he later returned and explained to me his belief that the future of personal styling would blend algorithmic recommendations with human taste. It sounded a lot like Stitch Fix, which would go public in 2017 at a $1.4 billion valuation while Amazon would go on to experiment with similar models before eventually abandoning them all.

At its core, Stitch Fix was trying to make personal styling accessible beyond the wealthy, combining algorithms with actual human stylists all while serving customers from the comfort of their own homes. It struck me as a model with the potential for real staying power, or at minimum a compelling acquisition target for a mainstream retailer or department store looking for new ways to reach customers and drive growth.

It didn’t quite work out that way. But the current CEO Matt Baer is confident a comeback is underway.