AI Overviews.
AI Overviews is the AI-generated summary Google shows above the search results page, synthesised from more than one source.
It does not replace the classic results; it sits above them and takes up much of the screen. The sources come from Google’s own search index, so being indexed is a precondition.
According to industry reports the share of queries it covers rose markedly towards 2026, but the distribution varies greatly by query type: very high for informational queries, very low for transactional and navigational ones.
Its effect on clicks runs both ways. On queries where the summary is shown, the click-through rate of the top positions falls; on the other hand pages cited inside the summary get a higher rate than rivals that are not. The losing side, then, is the page that ranks well but is not quoted.
The unit of selection is the passage, not the page. The summary takes a section from each of several pages and synthesises them; the whole page need not be relevant, it is enough for one section to answer the question fully.
Measuring it is hard, because the search console does not report impressions inside the summary as a separate line. The most reliable method is manual observation: searching the target queries regularly and recording who is mentioned.
The healthy response is to separate two things. Part of the traffic from informational queries will fall permanently, and that is not reversible. Being mentioned inside the summary, on the other hand, is a new visibility channel and not yet crowded.
The practical strategy follows from that: write informational content to be quoted and to have the brand mentioned rather than to earn a click, and move the conversion expectation to the query types the summary does not cover.
- Informational queries — highest coverage, largest effect.
- Comparison queries — high coverage; the summary describes both sides.
- Transactional queries — low coverage; commercial results are preserved.
- Navigational queries — very low coverage.
This distribution also determines how much AI Overviews will affect a given site: a publication whose traffic comes mostly from informational queries is affected far more than a shop living on transactional ones.
There is a difficulty on the measurement side that has to be known: the search console does not report impressions inside the summary as a separate line. The effect is read indirectly — if impressions stay flat while the click-through rate falls, a summary may be showing on those queries. The more reliable method is manual observation: searching the target queries regularly in a private window and recording who is mentioned.