Your search traffic report is about to tell a story that is not true. When an AI answer sits at the top of the results, fewer people click, even when they are just as interested in what you sell. If leadership reads visits as demand, it will see a decline that may not exist and fund the wrong fix.
The evidence is now hard to argue with. Digital Content Next's review of the research, published October 5, pulls together controlled experiments, observed browsing, and traffic data from thousands of sites. In a randomized field experiment with more than 1,000 Google users, researchers at the Indian School of Business and Carnegie Mellon University found outbound organic clicks fell 39.8% when an AI summary appeared, and searches that ended without any click rose 34.5%.
Other methods land in the same place. Pew Research Center's analysis of real searches by U.S. adults found people clicked a traditional result on 8% of visits when an AI summary appeared, compared with 15% when one did not. Ahrefs estimates that the top-ranked result loses 58% of its clicks when an AI summary sits above it.
Traffic coming back from AI assistants is growing, but it is small. In Chartbeat data from 2,576 publisher sites cited by the Reuters Institute, ChatGPT accounted for 0.02% of referrals. It is not a replacement for the clicks that are going away.
Ad buyers already see it. The IAB's September outlook found 86% of brand and agency buyers are changing, or expect to change, how they measure media because of AI tools and agents, and some are already using branded search and direct traffic as proxies. Their top measurement challenge is comparing AI-driven and traditional customer journeys.
The visit was never the point
For most companies, a search visit was a stand-in for something harder to count: a buyer who had a problem, found you, and started forming an opinion. AI answers now do part of that research before anyone reaches your site. Many of those buyers still arrive later, through a branded search, a direct visit, a sales conversation, or a referral. They just never show up as a click on the original question.
If visits stay the headline number, three things go wrong:
- Teams chase traffic back with more content that AI answers will also summarize.
- Budget drifts away from programs that are still shaping buyers who never click.
- The pages that do earn visits get judged on volume, not on whether they convert.
Move the scorecard to demand you can count
- Lead with outcomes. Qualified pipeline, sales-accepted opportunities, and revenue go first on the marketing scorecard. Visits move to a supporting line.
- Track demand that skips the click. Watch branded search volume, direct and returning visits, demo and contact requests, and how often sales hears that a buyer found you through an AI assistant.
- Ask buyers how they found you. A required "how did you hear about us" question on high-intent forms catches what analytics cannot.
- Judge pages on conversion, not reach. Fewer visitors means each visit is worth more. Measure the conversion rate of the pages that still earn traffic and fix those first.
- Reset the baseline before planning season. Compare this year to last with AI answers in mind, so a click decline is not read as a demand decline in 2027 budgets.
Three questions for the next leadership review
- If search clicks fell 30% next quarter, would we know whether demand fell too?
- Which number on our dashboard shows buyers choosing us, not just visiting us?
- Are the pages that still get visits built to convert buyers who arrive already informed?
This week
Put last year's search clicks next to branded search, direct visits, and qualified pipeline on one page. If clicks fell while the other three held, you do not have a demand problem. You have a reporting problem, and it is cheaper to fix.
