Fewer clicks: what should you measure alongside AI mentions?

What does the research show?

Pew's US browsing study observed fewer traditional-result clicks when AI summaries appeared. It is an observation, not a causal estimate for Hungarian tourism.

A place may be mentioned before someone later searches directly, opens a map or never visits its website. These journeys cannot always be attributed precisely to one AI answer.

Keep three views separate

• Observed AI mentions and factual accuracy.
• Measured website visits and known referral sources.
• Enquiries, bookings and their demonstrable origin.

Do not automatically label unknown traffic as AI success. In monthly reviews, distinguish direct measurement, estimates and unknowns. That makes progress discussions useful.

Sources and background

• Pew Research Center: Google users and AI summaries · 2025-07-22

Updated: September 17, 2026