Do AI Overviews Reduce Website Clicks?
This is a genuinely unsettled question, and we're not going to manufacture a precise-sounding statistic to fill this page. Published third-party studies show a range of effects — some finding significant click-through reduction on queries where an AI Overview fully answers the question, others finding smaller or more query-dependent effects. Below is what's currently known, our own methodology for tracking this across our clients' sites, and how we'll update this page as we gather more data.
Why we're not publishing an invented topline number
A page titled with a specific study name and a precise percentage is exactly the kind of unverifiable claim we've criticized elsewhere on this site — the entire trust argument for our business collapses if we do the same thing ourselves the moment it's convenient. What follows is honest instead: what's publicly documented, and our transparent, ongoing methodology.
What's publicly known
Third-party research on AI Overview click-through impact has generally found: - Effects vary significantly by query type — informational queries with a clean, complete answer tend to see larger click-through reductions than queries requiring more nuanced or comparative information - Effects vary by industry and content type - The picture is still evolving as Google adjusts how and when AI Overviews appear
Given how quickly this shifts, we'd point you to current third-party reporting for the latest topline figures rather than cite a specific number here that may already be stale by the time you're reading this.
Our own methodology (in progress)
We're tracking, across our own client base with their permission: - Search Console click-through rate changes on queries where an AI Overview appears vs. queries where it doesn't, for comparable query types - Whether being cited within an AI Overview correlates with any click-through recovery, compared to not being cited at all - How this varies across the SEO/marketing-services category specifically, given that's the vertical we have the most direct visibility into
What we'll publish once we have enough data
A genuinely dated, sourced analysis with our actual sample size and methodology stated plainly — including the limitations of drawing conclusions from one industry vertical. We'd rather publish a smaller, honest finding later than a large, invented one now.
In the meantime
See our full guide to GEO for the practical structural changes worth making regardless of exactly how large the click-through effect turns out to be — the actions (answer-first structure, schema, entity consistency) are useful whether the effect is large or modest.
Read our GEO guide for what to do regardless →
Written technical findings by senior strategists. Published transparent pricing.