Google Admits Search Console Can't Track AI Overviews Properly
Google's John Mueller admits Search Console flattens AI Overviews into one block, giving every link position 1 — while AI answers cut organic clicks by up to 38%.
Updated
The brief
- John Mueller confirmed Search Console treats entire AI Overviews as a single block, making traditional ranking tracking 'hard to do in a way that makes it useful'
- Every link inside an AI Overview is recorded as position 1, whether featured prominently or hidden behind a 'Show More' toggle
- Seer Interactive found AI Overview CTR fell 61%, and a field study showed AI Overviews cut clicks to organic results by roughly 38%
Google has admitted that Search Console cannot meaningfully track traditional rankings for AI features — and its reporting method gives every link inside an AI Overview the same top position, regardless of where it actually appears.
Search Advocate John Mueller made the admission in a Reddit thread, explaining that tracking traditional rankings for AI features is "hard to do in a way that makes it useful." He confirmed that Search Console treats an entire AI-generated overview as a single block, validating what many marketers have observed for months: forcing AI answers onto a 1-to-10 ranking scale no longer works.
The admission lands at a difficult moment for the industry. Clicks from organic search have declined for the past couple of years, and marketing teams face mounting pressure from clients and stakeholders to explain the drop or prove that traffic has simply shifted to AI. AI tool vendors have moved quickly into that vacuum, pushing narratives that suit their products rather than what measurement actually requires.
Why 'average position' misleads
The problem predates AI. Search Console's ranking figures rely on a simple mathematical average, and search results rarely cluster neatly around a middle value. Rankings shift depending on location, device and the result format Google chooses to display in the moment.
A page might rank position 1 whenever an AI Overview appears and position 19 in standard results. If both happen equally often, Search Console reports a tidy average of 10 — a number that matches what almost no real searcher saw. Changing views within Search Console for the same terms and pages produces different average positions depending on the cardinality and thresholds applied.
How block flattening works
The distortion deepens with how Search Console records AI Overviews. Instead of checking where an individual link sits inside a summary, Google ranks the overview box as a whole. Because AI boxes typically appear above organic listings, the entire box counts as position 1, and every link inside inherits that rank.
An eye-catching link card placed directly under the opening sentence gets position 1. A plain link buried in a "Show More" dropdown also gets position 1. To Google's database, both placements are identical, even though searchers are far more likely to click the first and overlook the second entirely.
Impression rules compound the problem
The flattening collides with impression-counting rules written before search results became interactive. An impression logs the moment a result loads, whether or not the user scrolls to see it — so all default links in an AI Overview count immediately, filling reports with views from people who read one sentence and left.
The rules invert for interactive elements. Links behind a "Show More" toggle log no impression until someone clicks to expand. Unread links in the main text get full credit for views they never had; useful links behind a toggle look nonexistent until deliberately opened.
The traffic impact is measurable
The reporting quirks distort metrics at scale. Seer Interactive found that AI Overview CTR fell by 61%, with click-through rates for cited websites dropping noticeably — partly because Search Console's automatic impressions skewed the math.
A separate field study covered by Search Engine Journal found that an AI Overview at the top of the page cut clicks to standard organic results by around 38%. AI Overviews now appear on 21% of searches, most often on informational queries where Google answers directly rather than sending users elsewhere.
Mueller said Google has tried to explain the rules in its help documents and asked the SEO community for practical ideas on how position could be measured when search pages no longer resemble a simple list. Modern results vary constantly: the same search can show three link cards on mobile, five on desktop, hide more behind a button, and change wording minutes later.
What to measure instead
The wave of AI visibility tools cannot patch this gap either, according to the analysis. A more reliable approach treats AI visibility as a binary question — whether the brand is cited at all — rather than chasing an average position of 1.2 that usually just means inclusion somewhere in the overview box.
Outcomes immune to impression logging — actual site visits, genuine inquiries and conversions — now tell marketers more than any ranking figure. Where page layout matters, inspecting the live screen directly or with automated tools reveals far more than Google's flattened data. The list of 10 blue links served the industry for 25 years; clinging to that model, the analysis concludes, will only cloud any understanding of the audience.
Based on Search Engine Journal
Filed under google-search-console, ai-overviews, google, seo-measurement, john-mueller
James Calloway
Show full bio
Market editor covering media and advertising at Marketing Herald.