Google confirms Search Console flattens AI Overviews into one block, so every cited link reports as position 1

A brand looking at its Search Console AI Overviews report to decide whether AI answers are helping or hurting has, according to Google Search Advocate John Mueller, been reading numbers the AI layout does not support. Dan Taylor at Search Engine Journal wrote up the exchange on 2026-09-24, and its practical consequence is that most of what Search Console reports for AI Overviews is either flat or automatic.

What Mueller confirmed

Mueller replied to a Reddit thread that tracking traditional rankings for AI features is "hard to do in a way that makes it useful", and that Search Console still treats entire AI-generated overviews as a single block. The two facts a brand can act on follow from that.

The first is block flattening. Rather than tracking where an individual link sits inside an AI summary, Google gives the overview box one overall rank. Because AI Overview boxes usually appear above the standard listings, the whole box is counted as position 1, and every link inside inherits that position in the report. Taylor's example: an eye-catching link card shown right under the first sentence and a plain link hidden inside a "Show More" dropdown are both recorded as position 1, even though a searcher is far more likely to click the first than the second.

The second is impressions. Under Search Console's normal rules, an impression is logged the moment a result loads onto the page, whether the searcher scrolls to see it or not. All the default links in an AI Overview count as impressions the moment the box appears. Links hidden behind a "Show More" toggle are the one exception: those do not count as an impression until someone deliberately expands the section.

The two things this breaks

Averages first. A page appearing at position 1 whenever an AI Overview appears and position 19 in normal results, at equal frequency, produces a Search Console Average Position of 10. Neither number was ever the searcher's experience. Taylor notes the same terms and pages can produce different average position numbers in Search Console when views change, because of the cardinality and thresholds Google applies to the report.

Impressions second. The Search Engine Journal piece cites SEJ coverage of a Seer Interactive analysis in which AI Overview click-through rate fell 61 percent, an effect Taylor attributes in part to Search Console recording so many automatic impressions that the ratio was skewed. A separate field study SEJ covered found that an AI Overview at the top of the page cut clicks to standard organic results by around 38 percent.

What to measure instead

Taylor's recommendation is to treat AI visibility as a yes-or-no question first, whether a brand is being cited in the box at all, then move to metrics that block flattening cannot distort: actual site visits, inquiries and conversions. Where the exact layout matters, checking the live SERP directly (or an automated tool that inspects the rendered page) reveals more than the flattened Search Console data.

The cost of not moving is a false sense of accuracy: a page reporting an average position of 1.2 in the AI Overview report almost always just means its link is included in the box somewhere, not that a searcher actually saw it near the top. Stakeholder education is the harder part of the work, and it starts by naming the two Search Console fields that no longer mean what they used to.

Source

• Search Console Uses Block Flattening For AIOs, Forget Position & Focus On Outcomes, Dan Taylor, Search Engine Journal, 2026-09-24