France launched AI Overviews on July 22, 2026, giving search researchers something rare: a clean, dateable before-and-after. Ahrefs tracked 963 domains using aggregated Google Search Console data across 28 days before launch and 9 days after. The headline finding: AI Overviews CTR (click-through rate, the share of searchers who click your link) damage is real, measurable, and highly uneven. The most-exposed sites lost 23.1% of their click-through rate in under two weeks. Sites with minimal exposure barely moved.
How large was the overall CTR drop across French search?
Across all 963 domains in the Ahrefs panel, average CTR fell 3.4% in the nine days following the July 22 launch. That figure sounds manageable until you look at the distribution underneath. The most AI Overview-exposed domains lost 23.1% of their CTR, roughly six times the panel average. And 82.2% of those highly exposed domains recorded a CTR decline, according to Ahrefs data published in 2026.
- Full panel average CTR change: -3.4%
- Highly exposed domains, CTR change: -23.1%
- Share of highly exposed domains that lost CTR: 82.2%
- Median CTR change across all domains: -5.7%
- Median click volume change: -8.8%
The -23.1% figure describes a minority of sites, Ahrefs notes. But the broader distribution is harder to dismiss. Nearly one in two domains in the study lost more than 10% of their daily clicks. One in twenty lost more than half, all within nine days of launch.
Is there a rule of thumb for estimating your own AI Overviews CTR risk?
Ahrefs ran a slope analysis across all 963 domains, comparing AI Overview exposure rate against CTR change. The result: a slope of -0.9. In plain terms, every additional percentage point of AI Overview exposure costs roughly one percentage point of click-through rate. If 15% of your queries trigger an AI Overview, expect to lose around 15% of your CTR. Ahrefs flags this as “crude but actionable”: pull your exposure rate from Site Explorer and you have your order of magnitude.
The current median AI Overview rate among studied domains is 5.4%, with 29% of domains already above 10%. Those numbers have moved in one direction in every market since 2024, according to Ahrefs.
Which sectors and domains took the hardest hit?
French-registered (.fr) domains fared worse than the panel average, losing 12.3% of clicks at the median versus 8.8% for all domains combined. The health sector stands out most sharply: health domains carry an AI Overview rate of 22.2%, the highest of any sector in the study, and their CTR dropped 20.4%. This aligns with earlier Ahrefs research finding that 44.1% of medical queries trigger an AI Overview.
The most extreme case in the French panel recorded +183% impressions alongside -1.9% clicks, more than tripling the page’s SERP (search results page) visibility while slashing the traffic it actually delivers. Rankings are holding. Clicks are not.
Does your industry matter more than the intent behind your queries?
Ahrefs argues intent matters more than vertical. A query answerable in a single paragraph is one you are likely to lose to an AI Overview. Queries that require a tool, an account, a price comparison, a purchase, or a human opinion hold up significantly better. Health queries are disproportionately affected because they are informational, produce short answers, and rarely involve a downstream transaction, making them ideal candidates for AI Overview absorption.
Is this a temporary shock or a structural shift?
Ahrefs’ global research suggests the impact deepens over time rather than stabilizing. In their earlier global study, the effect of AI Overviews on position-1 CTR measured -34.5% in spring 2025 and deepened to -58% by the end of 2025. Users adapt to not clicking. AI Overviews trigger on progressively more queries. Ahrefs states explicitly that nothing in their data suggests France escapes this trajectory, and they plan to re-run the analysis over a longer window to track the full arc.
What this means for AI-search visibility
Google is now answering users’ questions directly in search results before they ever click your link, and the French data confirms that the more of your content answers simple informational questions, the more traffic you lose, at roughly a one-for-one rate.
Think of it this way: imagine a bookshop that places a summary card in front of every book on the shelf. Customers read the card, get what they needed, and leave without opening the book. The book is still there, still on the shelf, and the shop even attracts more foot traffic than before. But purchases drop. That is what AI Overviews are doing to informational content right now, particularly in health, definitions, and how-to queries. The Ahrefs data makes this concrete: one French site gained 183% more impressions and lost 1.9% of its clicks at the same time.
The near-linear slope (-0.9) is worth pausing on. It suggests there is no safe threshold below which exposure stops mattering. It is a continuous cost, not a cliff edge. That has a planning implication: an AI Overview rate of 8% is not “low risk,” it is simply a smaller tax than 20%.
Hingewise’s read: the sites that appear to be holding up well despite high AI Overview exposure likely share a common characteristic worth examining further. Their query mix probably skews toward transactions, comparisons, or brand-driven searches that AI Overviews cannot fully resolve in a box. The next useful research step is segmenting the “resilient” cohort by intent mix rather than industry. Until that analysis exists, intent distribution is the most actionable lever available. The global trajectory (position-1 CTR halved over the course of 2025, and still moving) suggests that brands treating this as a temporary adjustment period are likely underestimating the structural nature of the shift.
Five things to check before AI Overviews reach your market
- Pull your AI Overview exposure rate in Ahrefs Site Explorer using the SERP features filter. Your current rate is a direct proxy for your traffic risk level.
- Segment your keyword list by query intent: purely informational queries (high risk) versus transactional, navigational, or opinion-based queries (more resilient).
- Identify the top 20 traffic-driving pages that answer informational questions answerable in one paragraph. Those are your highest-exposure pages.
- Assess whether those pages can be supplemented with content formats that require user interaction: calculators, comparison tables, original data, or community input.
- Establish a clean CTR baseline by intent category now. The France study was valuable precisely because Ahrefs had 28 days of pre-launch data. Document your numbers before any launch date is announced.
The France study is the cleanest natural experiment the industry has had on AI Overviews’ CTR impact, and the results are consistent with what global data already suggested. The most useful follow-up will be Ahrefs’ own re-run of this analysis over a longer post-launch window, which should clarify whether the -23.1% for high-exposure sites is a floor or a starting point.
Source: Ahrefs Blog, “AI Overviews France Impact” (2026)
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.