On Shopify's Q2 earnings call they stated that 75% of AI-attributed purchases in Q2 landed outside Shopify's top 100 categories.
AI agents are reshaping how niche products get found. Shopify's Q2 data shows 75% of AI-attributed purchases landed outside its top 100 categories — not because the mix shifted, but because structured product data lets agents match on precise intent instead of keyword popularity.
On Shopify's Q2 earnings call they stated that 75% of AI-attributed purchases in Q2 landed outside Shopify's top 100 categories. That might not necessarily be a surprise for some since Shopify's customer base skews niche.
At the same time, Shopify says its share of sales outside major categories has held steady since 2025. So the mix hasn't shifted, but how those sales are arriving has changed.
During the call Shopify's president illustrated the example of a car seat. When a shopper asks an agent for 'the best car seat that fits three across a sedan', a traditional search engine ranks on the keyword 'car seat'. An agent works through the dimensions, the vehicle type and the fact there are three seats and matches on all of it at once.
That's a significant shift because under keyword-ranked search, a niche product existed but usually stayed buried under bigger, better-optimised competitors because ranking rewards scale and ad spend nearly as much as relevance. A shopper had to wade through a ranked list built on popularity to find it, if they found it at all.
Commercial that's relevant because AI search powered by a structured catalogue converts at twice the rate of AI search working off a scraped website. Same shopper intent, same AI channel, the only variable is whether the underlying product data was actually built to be matched against. That gap is the difference between winning or losing a sale you didn't know you were competing for.
The underlying selection logic isn't platform-specific. Any agent matching on structured attributes will favour the more precisely-described product over the more prominent one, regardless of which platform hosts it. What differs right now is readiness because not every platform's merchants have a catalogue exposing those attributes the way Shopify's does.
The brands that work this out first won't necessarily be the biggest. They'll be the ones whose product data actually answers the question being asked.
That's the question MARK answers: does your product data speak the language agents are matching on. Build for Selection.
- —75% of Shopify's AI-attributed Q2 purchases landed outside its top 100 categories — the mix didn't shift, but how sales arrived did.
- —AI search powered by a structured catalogue converts at twice the rate of AI search working off a scraped website.
- —The winners won't be the biggest brands; they'll be the ones whose product data answers the precise question the agent is matching on.