What happened in France
AI Overviews went live in France on July 22, 2026. The most exposed domains lost 23.1% of their CTR after the launch, according to data from Ahrefs.
The study tracked 963 domains through Google Search Console, comparing 28 days before the rollout with 9 days after. The drop reflects click-through behaviour in search results, not disappearance from rankings. Pages can retain visibility while collecting fewer clicks.
For SEO this matters. Even when a page holds its position, its share of clicks may fall. Users see the answer in search and open the link less often. The query still exists, but the path to the site shifts.
How the impact was measured
Researchers compared 28 days before AI Overviews launched with 9 days after, using CTR as the primary metric — the ratio of clicks to impressions. That window lets you see the change in search behaviour right at the point when the interface altered.
The 23.1% figure applies to the most affected domains, not every site in France. A drop in CTR does not automatically mean a drop in rankings or impressions. A page can keep appearing in results but attract fewer clicks. When checking your own data, look at impressions, clicks and CTR together in Google Search Console.
Why the figure matters for SEO
A 23.1% CTR drop changes the value calculation for organic traffic. If your SEO model depends on clicks from search results, earlier expectations may now be too high.
For businesses this means checking actual data, not just positions. Ranking first no longer guarantees the same click volume. Users can get what they need on the search page and stop there.
Pages that used to collect clicks by offering short answers face higher risk. When a query can be closed directly in search, CTR for those pages is more vulnerable. The study does not extend this assessment to every query type, but the trend shows that the link between impression and click is weakening.
Short version: count visibility and traffic separately. That is the baseline check after AI Overviews launch.
What to check in Google Search Console
Start with the period around July 22, 2026. Compare data from 28 days before launch with 9 days after, if your project has comparable volume.
Look at overall CTR first. Then check whether clicks and impressions changed. If impressions stayed close but clicks fell, that looks more like a shift in result-page behaviour. If all metrics moved together, CTR alone will not be enough.
Next, split pages and queries into groups. Do not average across the whole site right away. An aggregate figure can hide domains or pages that took a harder hit. The Ahrefs study also isolates the most affected domains, so a project-wide number may look milder.
Capture the before-launch figures separately from the after-launch ones. That makes it easier to see the direction of change and not lose the baseline in later comparisons. For an SEO team this is a working record, not a one-off presentation metric.
How to interpret the CTR drop
A fall in click-through rate is not immediate proof of a content problem. When search adds a new way to answer a query, users may click through less often even when the page holds its old position.
Separate changes in the search result page from changes on your own site first. Check the period, compare clicks with impressions, and see which domains or page groups drove the drop. That order stops you wasting time reworking pages that lost clicks for an external reason.
For SEO assessment this is also a prompt to revise forecasts. CTR can no longer be treated as a stable quantity across all search conditions. It depends on how the user gets the answer and whether a reason remains to open the result.
One figure is not enough to decide on budget reallocation. The study shows a CTR drop for the most affected domains but does not describe every site and every query. The practical takeaway for a project owner is simple: measure your own trend in GSC before changing strategy.
What changes for site owners
AI Overviews add another risk factor for click volume to SEO. Positions still matter, but assessing results by rankings alone is no longer enough. A site can keep its search presence and lose clicks at the same time.
For reporting it makes sense to show at least CTR, impressions and clicks together, not one metric on its own. That way the team sees what actually shifted after launch. This is especially important for domains that depend on organic traffic and use clicks as the foundation of the funnel.
The France launch date offers a clean comparison point. Researchers took 28 days before it and studied 9 days after. You can use those values as a reference when reviewing your own data, keeping the same comparison logic.
The main conclusion is not about one country or one metric. A change in the search interface can cut CTR even where a site continues to get impressions. The 23.1% drop among the most affected domains shows the scale of the risk and requires checking actual clicks after AI Overviews roll out.
