A Shopper Asks Once, Not Twelve Times

Somewhere tonight, a shopper types a single question into an AI assistant instead of scrolling through ten blue links, and the assistant answers with a specific product from a store they have never heard of. No keyword stuffing got that product there. No page-one bidding war. Just a description good enough for a model to understand and recommend.

That shift is no longer just a vibe among marketers watching their analytics. Shopify put a number on it: AI-driven traffic and orders to its merchants' storefronts tripled year over year in the second quarter of 2026, and the company's overall revenue rose right along with it, to $3.6 billion, beating Wall Street's $3.4 billion forecast, a 36 percent increase from a year earlier.

The Traffic Tripled, but Search Did Not Shrink

The obvious assumption is that AI assistants are simply stealing sessions that used to come from Google or Bing. Shopify's own numbers say otherwise. Traditional search still accounts for roughly one-third of all sessions across its merchants' storefronts, and that traditional search traffic is still up over the past two years, not flat, not declining.

In other words, the AI channel appears to be sitting on top of the old one rather than eating into it. For a merchant, that reads less like a channel swap and more like a new, additive source of demand arriving alongside search traffic that was already there.

Finkelstein: A Complement, Not a Substitute

Shopify President Harley Finkelstein put the distinction in plain terms, describing AI search as "a complement to search, rather than a substitute for it." He added that the effect benefits smaller merchants particularly, the sellers who have historically struggled to outbid larger competitors for search visibility.

That framing matters because it is the opposite of what smaller players usually experience when a dominant discovery channel changes shape. Algorithm updates and paid-search inflation have tended to favor whoever can spend the most. Shopify's account of this shift points the other way.

Why Publishers Are Bleeding While Merchants Gain

Set this against the story that has dominated 2026 coverage of AI search so far, which is a publisher's story. News and media sites have reported measurable drops in click-through rates as AI-generated answer summaries satisfy a reader's question before they ever click through, cutting traffic and advertising revenue in the process. For publishers, AI search has largely been cannibalistic.

Shopify's data describes the mirror image for e-commerce. An AI shopping answer does not resolve the reader's need the way a summarized news answer does. It still has to end in a transaction, on a merchant's storefront, and Shopify's figures suggest that transaction is arriving in addition to the sale that traditional search was already generating, not instead of it.

75 Percent Outside the Top 100, and What EU/UK Sellers Should Do Now

The most striking figure in Shopify's release may be this one: 75 percent of AI-attributed purchases in the second quarter fell outside the platform's top 100 product categories. Traditional search ranking rewards popularity and existing link authority, which tends to bury niche and long-tail products under whatever is already selling well. AI recommendation, on this evidence, is doing the opposite, surfacing inventory that would rarely reach page one of a conventional search result.

For a small or mid-size e-commerce owner in the EU or UK, that is a concrete reason to start treating legibility to AI shopping agents the way they have long treated SEO: clean, structured, complete product data, accurate attributes, and descriptions written for a model to parse and recommend, not just for a human to skim past a hero image.

None of that should be mistaken for proof of a lasting trend. This is one vendor's self-reported data from a single quarter, not an independent audit, and not a guarantee that AI shopping agents will keep behaving this way as the market matures. Treat it as an early, concrete signal worth acting on cautiously, with modest investment and close monitoring, rather than as a settled fact to bet the catalog on.