AI Sends You Less Traffic. It's Worth More.

Every few weeks another chart makes the rounds: search clicks are shrinking, AI answers are swallowing the results page, the traffic is drying up. If you run a small business, the conclusion seems to write itself. Why chase visibility in AI if nobody clicks anymore?

The numbers behind that fear are real, and we've published them ourselves. But they're half the story, and the missing half is the one that matters to your bottom line: AI sends fewer visitors than classic search does, and the visitors it sends are worth more. Often several times more.

The pipe is narrowing. The water coming through it is getting cleaner.

The half you've heard: zero-click searches

The click really is shrinking. The industry calls these zero-click searches: a question asked and answered without a single visit to the open web. In the first four months of 2026, about 68% of US Google searches ended that way, up from roughly 60% in 2024 (SparkToro, 2026). When an AI summary appears above the results, people click a traditional result about 8% of the time, versus 15% without one (Pew, 2025). And AI answers have cut the click-through rate of the #1 result: one study measured a 34.5% drop (Ahrefs, 2025), and a follow-up a year later measured 58% (Ahrefs, 2026).

If the number of visits were the whole scoreboard, that would be grim. It isn't. A visit was never the goal; a customer was. Which raises the question that matters here: what are the visits that do arrive actually worth?

The half you haven't: what an AI visitor is worth

Three independent datasets now point the same direction. (In your own analytics, this segment shows up as AI referral traffic, sometimes labeled LLM traffic.)

The biggest one flipped. Adobe tracks AI-referred visits across more than a trillion visits to US retail sites. In May 2026, that traffic had grown 138% year over year, and more than 14x since Adobe started tracking it in late 2024. Early on, those visitors converted worse than everyone else, at roughly half the rate of non-AI traffic. A year later the same measurement had fully reversed: by May 2026, AI-referred visitors converted 54% better, up from 42% better just two months prior, and spent 53% more time on site once they arrived (Adobe, 2026; Digital Commerce 360, 2026). The channel didn't just grow. It matured, from tire-kickers to buyers.

The commerce platform agrees. Shopify's own storefront data for the second quarter of 2026 shows AI-referred visits growing about three times year over year, converting roughly 80% better than organic search visitors, and about twice as well in categories where people compare details before buying (Shopify, 2026). Notably, organic search grew too. Shopify's read is that the two channels are doing different jobs, not fighting over one.

And it's not just ecommerce. One agency published a case study of a single B2B company's traffic: visitors referred by AI assistants converted at anywhere from roughly two to nine times the rate of organic search visitors, depending on the assistant (Seer Interactive, 2025). The honest print on that one: it's one company, and AI referrals were a tiny fraction of its traffic. Treat it as a ceiling someone actually measured, not a promise.

Now the honest print on all of it, because these numbers deserve the same scrutiny we give everything else. Adobe's and Shopify's figures are their own first-party data; nobody outside can audit them, and both lean heavily on retail. None of this is an industry benchmark you can plug into a forecast. But three datasets, from three companies with three different vantage points, all moving the same direction, is the kind of signal worth acting on.

Why the AI visitor arrives warmer

There's no mystery in the mechanism. When someone researches through an AI assistant, the assistant does the middle of the buying journey for them: it searches, compares the options, answers the follow-up questions, and names a shortlist. (We walk through exactly how in how AI search really works.) All of that used to happen on your website, across five visits from someone "just looking."

The person who clicks through after that conversation has already done the comparing. They arrive closer to a decision, which is exactly what the conversion numbers show. Shopify sees it in where people land, too: about half of AI-referred visits skip the homepage entirely and land directly on a specific product page. The assistant isn't dropping people at your front door to wander. It's walking them to the exact shelf.

So the browsing didn't disappear. It moved inside the AI conversation. What comes out the other side is fewer visits, further along.

The catch: warm traffic only flows to the named

Here's the part that should change how you spend your time. Every number above describes the businesses the AI actually mentions. Warmer clicks are the prize for being in the answer. If the assistant recommends your competitor instead, you don't get fewer clicks. You get none, and no report tells you it happened.

That's why "traffic is dying" is the wrong lesson to take from the zero-click charts. The right lesson: each mention in an AI answer now carries more value than a blue link used to, because the visitor behind it is closer to buying. The game isn't defending click volume. It's being one of the businesses that gets named, and the way you earn that is the same foundational work that has always mattered, because AI search still runs on search.

See where you stand

The first step is knowing whether any of this warm traffic can currently reach you. Our AI Visibility Analysis runs 25 realistic customer questions across ChatGPT, Claude, Gemini, Perplexity, and Grok, with live web search on, and shows how often each one names your business, who it recommends instead, and where to focus first. If the assistants aren't mentioning you, the most valuable visitors on the web are being walked to someone else's shelf.

Key takeaways

• Search clicks are genuinely shrinking. That half of the story is true.
• The visits AI does send convert far better: 54% better in Adobe's US retail data, roughly 80% better on Shopify storefronts, and up to several times better in one measured B2B case.
• Adobe's data shows the channel maturing: AI visitors converted worse than average a year ago and better than average now.
• The mechanism is simple: the assistant does the comparison shopping, so the click that survives arrives closer to a decision.
• The caveats are real: first-party data, retail-heavy, no audited benchmark. The consistent direction across three independent sources is the signal.
• None of it helps a business the AI never names. Being in the answer is the new front door.