Shopify's AI-Driven Orders Tripled Last Quarter. 75% of Them Came From Outside the Top 100 Product Categories.
Shopify reported Q2 2026 earnings on August 5: $3.58 billion in revenue, up 34% year-over-year, $116 billion in GMV, and free cash flow margin near 18%, all beating analyst estimates. That's the headline every outlet led with, and it's a fine headline. The number I actually care about as someone who thinks about where solo e-commerce traffic comes from is buried further into the merchant data: 75% of AI-attributed purchases on Shopify came from outside the platform's top 100 product categories. If you've been optimizing for the same crowded keywords and ad categories as everyone else, the fastest-growing discovery channel appears to be finding the niche stuff instead.
The numbers, in order of how much they matter to you
Start with scale: AI-driven traffic and AI-driven orders to Shopify stores each tripled year-over-year. Shopify's Sidekick AI assistant, the tool merchants use to manage their own stores, handled roughly 34 million merchant conversations in the quarter, with daily active merchant usage up 3.6x year-over-year. Those numbers describe genuine, fast-growing adoption of AI both on the shopping side (customers using AI tools to find and buy products) and the merchant side (store owners using AI to run their operations).
Then the category detail: 75% of AI-attributed purchases came from outside Shopify's top 100 product categories. That's the number worth sitting with, because it cuts against the intuitive assumption that AI shopping tools would just accelerate demand for whatever's already popular. Instead, the data suggests AI search and shopping assistants are disproportionately surfacing long-tail and niche products, the ones a traditional Google search or a paid ad campaign in a crowded category would have buried under bigger, better-funded competitors.
Why this makes sense if you think about how AI shopping actually works
A traditional product search optimizes for relevance within a category people already know to search for. If you sell something specific, a particular kind of ergonomic keyboard accessory, a niche supplement formulation, a specialized tool for a hobby most people haven't heard of, you're competing against every other seller in a category defined by the search term a customer typed. An AI shopping assistant works differently: a customer describes a need or a problem in their own words, and the assistant matches that description against product data rather than against a fixed category taxonomy. That kind of matching is structurally better at surfacing something specific and unusual than a keyword search is, because it doesn't require the customer to already know the exact term for what they want.
If that's actually what's happening here, and the category-distribution number is consistent with it, then AI shopping discovery isn't a scaled-up version of existing search behavior. It's a different discovery mechanism that rewards specificity over category dominance.
What this means concretely if you run a niche Shopify store
If your product sits in a crowded, obvious category (generic phone cases, generic supplements, anything where the search term is one everyone already uses), this data doesn't obviously help you yet; you're likely still competing the same way you always have. But if you sell something specific and less commonly searched, this is a real signal that AI shopping assistants may already be finding your product through a path that has nothing to do with your SEO ranking or ad spend in that category.
The practical move is making sure your product data is legible to an AI assistant doing that kind of matching, not just to a human scanning a category page. That means detailed, honest product descriptions that describe the actual problem your product solves and who it's for, structured data (proper schema markup, accurate attributes) that a model can parse cleanly, and titles that describe function rather than relying on brand jargon a customer wouldn't type into a chat assistant. None of this is new advice in the abstract, but the AI-attribution data gives it a sharper, more urgent reason to actually do it now rather than treating it as a someday task.
The honest caveat
This is Shopify's own reported data about its own platform, not an independent third-party audit. "AI-attributed" purchase tracking is a self-defined, still-emerging metric across the entire e-commerce industry, and different platforms measure attribution differently, which makes cross-platform comparison unreliable and makes even the underlying methodology here something outside parties can't fully verify. Treat the direction (AI discovery skewing toward niche categories) as a real and plausible signal worth acting on. Treat the specific 75% figure as directionally useful, not as a precise, audited statistic you should build a financial model around.
What I'd actually do
If you run a Shopify store selling something specific rather than something generic, spend an hour this week rewriting your top three product descriptions as if you were explaining the product to someone describing their problem in plain language, not searching a category. Add or verify structured data on those listings. Then watch your traffic sources over the next month for any AI-assistant referral pattern you weren't tracking before. This costs an afternoon and touches nothing about your existing SEO or ad strategy; it just adds a second discovery path that this data suggests is already growing faster than the traditional ones.
Author
Lukas
@lukcombinator