Search Console data has been inflated, which marketers say has made reporting to clients difficult.
Google is fixing a long running bug on Search Console, but its admission of the error is raising questions over transparency for marketers.
The bug, which inflated impression counts, was reported as a logging error which caused the tool to over-report impressions from May 13, 2025 onwards.
Google wrote on its data anomalies Search Console support page: “A logging error is preventing Search Console from accurately reporting impressions from May 13, 2025 onward.
“This issue will be resolved over the next few weeks; as a result, you may notice a decrease in impressions in the Search Console performance report. Clicks and other metrics were not affected by the error, and this issue affected data logging only.”
The tech giant reported that two days will be missing from the bulk data report and will not be recoverable for some properties (February 28, March 1, 2026).
Google claims this bug fix will be a chance for SEO teams to access more specific insights into their performance. But, not everyone believes it is being transparent about the impact these metrics may have had on click-through rates or overall clicks.
“I welcome the correction, but here’s why Google isn’t being fully transparent,” explained head of SEO and AI search at Adorama, Elie Berreby.
‘Categorically false and deeply misleading’
The impressions were at least partially inflated due to Google logging third-party scrapers like SerpApi, which OpenAI uses to trawl through the Alphabet-owned company’s organic search engine results to help train its LLM.
That’s not all. “There’s also a high probability that Google Search counted AI overviews containing links as impressions (and we know these rarely lead to actual clicks),” Berreby said.
Google’s statement insists that clicks and other metrics were not affected and it was simply a “data logging” error, which Berreby called “categorically false and deeply misleading”.
He said if impressions are inflated, it’s mathematically impossible for clicks to remain the same without the click-through rate (CTR) taking a huge hit since the CTR is simply a calculation of the number clicks divided by the number of impressions (x100).
Why now?
There is some public speculation on why Google has decided to now announce that impressions have been inflated for an extended period.
Berreby suggested in a LinkedIn post that Google has only just figured out how to filter synthetic impressions, so they were reluctant to address the issue before they could offer a fix.
“The information at least now is transparent, but it should have not taken almost a year for them to come out with it,” Peter Rota, senior technical SEO manager at HUB International, told Affiliate Leaders.
“Additonally, they are 100% wrong about it not affecting other metrics. CTR percentage and average position are based on impressions. If you’re over reporting impressions that will affect CTR and also likely where you’re being ranked.”
Metrics aren’t measured in a vacuum. Each one feeds into another, and to isolate one would be to ignore the correlation and causation between them.
“Over the last year we have seen a significant uplift in impressions, which we had initially put down to the introduction of AI Overviews and an increase in zero-click searches,” explained James Pearce, senior SEO specialist at Receptional.
“Now we know these weren’t genuine impressions, it makes reporting for our clients incredibly difficult to quantify, with a reduction in CTRs despite improvement in total clicks. This means SEO teams have potentially been focusing on the wrong areas of their sites to improve.”
Phantom impressions
As it stands, links visible within Google’s AI Overviews count as a regular impression, but brand mentions aren’t counted. This is problematic, Berreby explained, because Google refuses to share AI Overviews data “for an obvious reason”.
“Google designed AI Overviews to provide immediate superficial answers. This naturally
reduces the likelihood that a user will click through to a website. Financially, this has hurt
websites but benefited Google.”
This keeps users engaged within the Google ecosystem for longer – if users never click through to a site, then they aren’t being sent through to a publisher or brand, where they are exposed to other ads, meaning less revenue for Google.
AI Overviews containing links are being counted by Google as standard impressions.
“Google is logging a simple view as an impression, but this type of impression is less likely to
generate a click. And this alters a website’s metrics. It artificially inflates top-of-funnel
visibility while decreasing CTR (again). Because of this, organic search might look less
effective than it actually is,” Berreby continued.
Data pollution
Impressions and CTRs are often used by marketers to evaluate the effectiveness of their titles and meta descriptions, and at an executive level, impressions can be used to estimate organic search market share and brand visibility.
“When a marketing team sees their CTR tanking, they might spend countless hours rewriting
their meta data, running A/B tests, or even completely pivoting their content strategy to fix a
problem that doesn’t actually exist,” Berreby said.
The skewed CTR has led to data pollution, as the metrics cannot be trusted, and budget may have been allocated to a ‘phantom’ audience, which may not ever materialise to paying customers.
The fix
Where do marketing teams go from here? If their own measurements are being inflated by the search engine, what can they do to get accurate metrics?
“The short answer is simple: nobody apart from Google itself can fix these data anomalies. Until Google fixes how impressions are counted, clicks will be the most reliable indicator,” Berreby added.
With both the AI Overviews stealing engagement, making clicks less likely, but still counted as ‘impressions’, as well as LLMs scraping Google’s organic search results creating synthetic impressions, there is no easy resolution.
“Two different causes, both impacting impressions, made two key performance indicators, CTR and average position, deeply inaccurate. That’s why after about 11 months of denying an obvious reality, Google had to admit it.
“But instead of being transparent, Google tried to minimise the problem and claimed that
clicks and other metrics were not affected by the error.
“When data drives entire strategies, transparency is the bare minimum. Google needs to be far more accurate in its public statements, and the search industry must set a standard of critical thinking where we analyse anomalies instead of celebrating them,” Berreby concluded.
