Collison (Stripe) on the e-commerce revolution: AI agents will shop instead of people

The internet was created for shopping. But while e-commerce previously worked through advertising, algorithm recommendations, and SEO, AI agents are rewriting all the rules of the game. Very soon machines will make purchases instead of people — and this will transform the entire online retail industry into a completely different ecosystem. This was explained by John Collison, co-founder of the payment platform Stripe, which over the past ten years has witnessed all the metamorphoses of e-commerce from the inside.

Keyword search is yesterday

One of Collison's key ideas sounds radical: keyword search as a way to find a product is a ridiculous mechanism, a relic of the past. When you type "red wool sweater XL" into the search bar, you're doing exactly what you would have done in 2005. But AI agents will work completely differently. An agent will analyze the context: what's your clothing style, budget, when you need the item, for what occasion. Based on this, it will select the optimal option — often without any human involvement at all. This is already happening on Netflix or Spotify, when the service picks a movie or track for you without your input. Soon this logic will take over all of e-commerce.

Brands prepare to speak to machines

The harshest change: if an agent is your primary customer, you need to convince it not with beautiful packaging or an emotional brand story, but with logic and data. Brands and sellers will have to describe products not for humans, but for algorithms. This is similar to SEO, but an order of magnitude more complex and critical. Previously, a marketer wrote for humans who wanted to see a beautiful picture and a touching story. Now you need to describe a product in a way that a machine understands its characteristics, compares it with alternatives, and chooses it.

"Brands are used to appealing to human taste and emotions.

Now they need to learn how to appeal to machines," Collison said in an interview with Bloomberg.

Can a machine copy taste?

The main question: can an AI agent truly copy human taste? Collison believes yes — but only if the agent is loaded with the right data about the user's preferences. If you always wear minimalist clothing in calm colors, the agent will learn this and will select items in this style, even if the new product is colorful. But this raises other questions:

• How will brands compete if the judge is an algorithm?
• What will happen to marketing when a machine chooses?
• Will brands have a chance to surprise or seduce the buyer?

What does this mean

This is not just a technological shift — it's a complete upheaval in online retail. E-commerce will stop being about how to show a product to a person more beautifully and vibrantly, and will become about how to describe it correctly so that a machine chooses it for its owner. Companies that adapt to this first will gain a huge advantage. And those who wait risk being left out of this transformation.