Google’s own Search Console AI search reporting counts impressions that a user may never have seen, assigns positions that do not reflect where a link sits inside an AI Overview, and rolls AI data into the same view as regular web search. The report became globally accessible on August 31, 2026. Google’s John Mueller has since confirmed: the current approach does not map well to how AI search actually works, and Google does not have a better solution yet.
What does the new Search Console AI search report actually show?
The report, announced in June 2026 and rolled out globally on August 31, 2026 according to Search Engine Journal, focuses primarily on impressions: how many times a URL appeared in AI search surfaces, including AI Overviews and AI Mode (Google’s conversational search experience). The data is filtered, meaning the numbers shown are already included in the regular Web search performance report. Adding both figures together would double-count the same traffic.
Why do these three metrics specifically mislead?
A Reddit post laid out the problems clearly, and Google’s John Mueller agreed with the assessment. There are three distinct ways the current reporting can give site owners an inaccurate read on their actual AI search presence, as summarized by Search Engine Journal.
- Impressions count links the user never saw. Google’s standard impression rule counts a link as “seen” when the page of results is served, not when the user actually scrolls to it. If an AI Overview rendered and your link appeared in it, that is logged as an impression even if the user never scrolled far enough to see it. Visibility is overstated.
- “Show More” links are undercounted. Links hidden behind the “Show More” expansion in an AI Overview are not counted until a user clicks to expand them. This goes the other way: if your site appears in that collapsed section, its actual exposure is understated in the report. That is meaningful information a site owner never sees.
- Position does not measure your link, it measures the AI block. The position metric in the Gen-AI report tracks where the AI Overview feature appeared on the search results page as a whole unit, not where an individual link sat within the Overview itself. For SEO purposes, the position of the link inside the feature is what matters. The current metric does not provide that.
What did Google’s John Mueller confirm?
Mueller responded to the Reddit discussion and agreed with the Redditor’s reading. He wrote, according to Search Engine Journal: “This is pretty much it. We tried to document it as clearly as possible in the help center page. Position for these is hard to do in a way that makes it useful, so we’re currently tracking it like we do for many search features (as a block), and it’s not separated out in the Gen-AI performance report.”
He also acknowledged the broader tension: “Search results pages have a lot of ways for users to interact nowadays, so the old ‘position 1-10’ is hard to map, or to make useful for site owners. If any of you have thoughts on what would be useful in terms of tracking position, I’d love to hear and am happy to discuss with the team.”
The framing is honest. Mueller is not dismissing the concern. He is saying the problem is genuinely hard and that Google is still working out what useful measurement would even look like for AI search surfaces.
Before drawing conclusions from your Search Console AI data
- Do not add the Gen-AI report figures to your regular Web search figures. The AI data is already inside the Web search total.
- Treat impression counts as reach estimates, not confirmed views. A user may not have scrolled to the AI Overview at all.
- Check whether your site appears in “Show More” sections manually by searching your target queries. The report may be undercounting those appearances.
- Do not interpret the “position” figure as where your link ranked inside an AI Overview. It reflects only where the AI Overview block itself appeared on the page.
- Compare trends over time rather than relying on absolute numbers. Relative movement is more reliable than the raw figures given the current measurement limits.
- Cross-reference with click data: if impressions are rising but clicks are flat or falling, that can indicate your link is appearing in non-visible positions within AI Overviews more often.
What this means for AI-search visibility
Here is the clearest way to think about this: if you checked Search Console today and saw strong AI Overviews impression numbers, you might feel your brand is visible in AI search. But those numbers could be inflated by links no one saw, deflated by links hidden behind “Show More,” and reporting a position that has nothing to do with your actual placement. The dashboard looks familiar, but it is measuring a different thing than what it appears to measure.
That gap matters more than it might seem. Most brands and SEO teams still treat Search Console as the primary source of truth for search performance. When the tool applies ten-blue-link (the traditional list of ten organic search results) era logic to AI search surfaces, it produces numbers that look like the old metrics but represent something meaningfully different. The risk is not that the data is wrong. The risk is that it looks right enough to act on without questioning it.
What the sources do not say, but is worth noting as a Hingewise observation: this situation creates an asymmetry between brands that interrogate their data and brands that accept it at face value. A site owner who reads a high impression count and concludes “we have good AI visibility” may be making decisions on overinflated figures. A competitor who notices the Show More understating problem may discover untapped AI presence they did not know existed.
Mueller’s invitation for community input on what better position tracking would look like is also a signal worth reading carefully. Google is effectively saying the measurement framework for AI search is still open. That is unusual. It suggests the industry will likely see updated reporting logic as AI search surfaces mature. For now, any metric coming out of the Gen-AI performance report should be treated as directional, not definitive.
The metric to watch going forward is not the absolute impression count but the gap between impressions and clicks inside AI Overviews. If that gap widens over time, it raises a legitimate question about whether appearing in AI Overviews is driving meaningful traffic or simply fulfilling the query directly, keeping users on the search results page.
Author: Lam Nguyen, Hingewise
Sources:
Search Engine Journal, “Google Admits Search Console Reporting For AI Search Is Inadequate,” September 2026. https://www.searchenginejournal.com/google-admits-search-console-reporting-for-ai-search-is-inadequate/589236/
This article was drafted with AI assistance from the sources cited above and checked against them before publishing. The "What this means for AI-search visibility" section is Hingewise's editorial analysis, built on the frameworks and positions we use in our GEO work. How we work: Editorial policy.