The “all clicks eliminated” framing overstates both where AI Overviews actually appear and what Google has said about their design intent. Google’s own documentation on AI Overviews describes them as appearing primarily for a subset of query types, generally more complex, exploratory, or multi-part informational queries where synthesizing information from multiple sources adds genuine value, not as a universal replacement layer over all search results. Google has also stated the design explicitly includes links to sources within and alongside the AI-generated overview, specifically framed as preserving the opportunity for users to click through, which directly cuts against the premise that the feature is designed to fully eliminate click-through as a category.
Why complex and transactional queries retain click-through behavior
The core mechanism worth understanding is that not all queries carry the same relationship between “getting an answer” and “needing to click somewhere.” A simple factual query, a unit conversion, a basic definition, a single verifiable fact, has always been more exposed to full satisfaction within the search results page itself, a pattern that predates AI Overviews entirely and showed up with earlier direct-answer features like featured snippets and calculator boxes. AI Overviews extend that same dynamic to a wider range of queries by synthesizing multi-source answers, but the underlying principle, that simple, single-fact intent is more susceptible to on-page satisfaction than complex intent, isn’t new or unique to AI.
Where this dynamic breaks down is precisely the query territory the original assumption tends to ignore: queries involving genuine comparison, verification, transaction, or trust-building. A user comparing products before a purchase decision typically wants to see the actual product pages, read genuine reviews, compare specific details, and ultimately transact somewhere, needs an overview summary structurally can’t fully satisfy regardless of how well-synthesized it is, since the endpoint of that query is an action (a purchase, a booking, a sign-up) that has to happen on a specific site. A user researching a YMYL topic, a health concern, a legal question, a major financial decision, has a well-established, persistent behavioral pattern of wanting to verify information against a primary, trusted, specific source rather than stopping at a synthesized summary, precisely because the stakes of getting it wrong are higher for these topics. This isn’t a hopeful assumption about user behavior, it reflects long-documented patterns in how people engage with search results for high-consideration or high-stakes queries, patterns that existed well before AI Overviews and that the underlying complexity/stakes of the query, not the mere presence of a summary above the results, continues to drive.
What Google has actually said about design intent
Google’s own public documentation and communication around AI Overviews has repeatedly emphasized that the feature is intended to work alongside, not instead of, traditional links to web content, explicitly citing the inclusion of source links as part of the feature’s design. This is a meaningfully different stated intent than “replace the need to visit websites,” and it’s worth treating this stated design goal as a real, citable position, while still being honest that stated intent and observed aggregate outcome aren’t automatically the same thing; independent, verifiable data on exactly how AI Overviews affect aggregate click volume across all query types is still an evolving, actively studied area, and no single, settled, universally-agreed figure exists for the actual magnitude of any click-through change, so any specific percentage claim about “clicks eliminated” circulating in industry discussion should be treated as, at best, a third-party estimate tied to a specific dataset and time period, not a fixed, Google-confirmed fact.
Practical implication for SEO strategy
The practical response to the “AI Overviews will eliminate all clicks” framing isn’t to dismiss the real, genuine pressure AI Overviews put on simple, single-fact query traffic, that pressure is real and worth planning around, it’s to recognize that the risk is concentrated, not universal. Auditing a site’s actual query portfolio for where genuine risk concentrates (simple factual, single-answer-satisfiable queries) versus where click-through behavior is more likely to persist (comparison, transactional, YMYL-verification, and generally complex or multi-step queries) gives a much more actionable picture than treating the whole organic traffic base as equally exposed.
For content strategy specifically, this means recognizing that content built around simple, easily-summarized single facts is more vulnerable to full on-SERP satisfaction, while content built around genuine comparison, expert analysis, personalized guidance, or content that supports a transactional or trust-verification step retains a more durable reason for users to click through even when an AI Overview is present. Rather than treating the entire organic channel as facing uniform, existential pressure, the more defensible planning approach is triaging by query complexity and intent type, reinforcing the kinds of content and query targets where the persistent, well-established click-through behaviors described above are most likely to continue holding, while treating the simple-fact end of the query spectrum as the genuinely higher-risk category deserving separate strategic attention.
