AI-driven traffic and orders to Shopify stores tripled year-over-year in the second quarter. That’s the number that stood out on the company’s earnings call, and TechCrunch AI reports it’s a big part of why Shopify beat Wall Street’s forecasts. President Harley Finkelstein put it plainly: AI has become a “complement to search, rather than a substitute for it.”
This matters because it runs against the story we’ve been hearing everywhere else. In online publishing, AI summaries have cut click-through rates, dragging down traffic and the ad revenue that depends on it. Shopify is describing the opposite effect. AI isn’t eating its channel. It’s opening a new one.
📊 The numbers behind the beat
Shopify’s Q2 results were strong across the board, according to TechCrunch AI:
- Revenue rose 36% to $3.6 billion, past the $3.4 billion analysts expected.
- Gross operating profit climbed 31% to $1.71 billion, ahead of the $1.63 billion forecast.
- AI-referred traffic and orders tripled from a year earlier.
- Traditional search sessions are still up 1.3x over two years, holding roughly a third of all storefront sessions.
That last point is the one to sit with. AI growth didn’t come at search’s expense. “Search remains one of our largest sources of buyer traffic to our merchants, and it’s still growing,” Finkelstein told analysts. Both channels are expanding at the same time.
🛒 Why AI search works differently for shopping
Finkelstein gave analysts a clear explanation, and it’s worth understanding. Traditional search engines rank pages by popularity against a handful of keywords. AI agents work differently. They make multiple calls into Shopify’s catalog, using richer structured data to match products to what a buyer actually wants.
His example: a shopper asks an AI assistant for the best car seat that fits three across a sedan. Keyword search fixates on “car seat.” An agent understands the real need, the dimensions, the vehicle type, and the requirement for three, then searches across all those constraints at once. It finds the product that works, not the one that happens to rank highest.
That precision shows up in the behavior data. Half of all AI-referred sessions land directly on a product page, which Shopify says is 2.5 times more than traditional search delivers. The shopping journey is getting compressed, and buyers arrive closer to a purchase.
🎯 The long tail is the real winner
Here’s what I find most telling. Shopify said 75% of AI-attributed purchases in Q2 happened outside its top 100 categories, the zone it calls its “sweet spot.” AI is surfacing products that keyword search tends to bury, and that helps the smaller merchants who make up most of Shopify’s customer base.
Think about what that means. Popularity-based ranking rewards the stores that are already big. Intent-based agents can match a niche product to a specific need, giving long-tail sellers a fairer shot at being found. For a platform built on millions of small merchants, that’s a structural advantage, not a one-quarter bump.
🔌 Building for the agent era
Shopify isn’t just watching this happen. The company has built connectors to Claude, ChatGPT, Perplexity, Manus, Replit, and Vercel, plus vibe-coding platforms like Lovable that let merchants build on Shopify however they want. It’s also leaning on its trusted checkout as AI takes a bigger role in transactions, betting that when an agent completes a purchase, buyers will want a checkout they already trust.
What to watch next
One quarter doesn’t settle the debate over how AI reshapes commerce. But Shopify just offered the clearest counterexample yet to the “AI kills traffic” narrative, backed by real revenue and behavior data. The takeaway for practitioners: structured product data and agent-ready catalogs are becoming a competitive edge, not a nice-to-have.
If you sell online, the signal is to make your catalog legible to agents now, while the shift is early. Expect more platforms to report their own AI-referral numbers in the coming quarters, and expect “agent-ready” to become a standard line item in e-commerce strategy. You can find the full breakdown at the original source.