AI Overviews France: Which E-commerce Sectors Lose 23.1% CTR
Summarize this article with AI
23.1% CTR in nine days: the number that changes the game
23.1%. Nine days. That’s the number I take away from the Ahrefs study published on September 16, 2026. On July 22, 2026, France became the last major market to receive AI Overviews. For once, we could observe the before and after without guessing.
963 domains tracked in Google Search Console. 28 days before launch. 9 days after. Ahrefs’ data team, led by Xibeijia Guan, isolated the AI Overview effect. The data from 963 French domains is clear. This time, no algorithmic noise blurs the reading, and the launch has a date.
The most exposed domains lost 23.1% of their CTR in nine days. In nine days, not over a quarter. The comparison with low-exposure domains is stark: they barely moved. The cause is identified. It’s the AI Overview, not a classic algorithm update.
This loss doesn’t hit a handful of unlucky sites. 82.2% of highly exposed domains lost CTR. At that level of exposure, getting through unscathed becomes the exception.
Across the entire panel, CTR declined by 3.4% in nine days. For the most exposed domains, the drop is six times stronger. The gap is clear: exposure to AI Overview is a measurable risk factor, not a vague concept.
I look at 15 sites per week. I already see this number in the Search Console data of my e-commerce clients. Product pages that keep their position. Buying guides that lose clicks. The question isn’t « Is this happening? » It’s « On which pages, exactly? »
The Ahrefs study goes further: it gives a method to estimate your own exposure. That’s what makes this work useful for an e-commerce owner. We measure instead of panicking.
For many e-commerce owners, this launch was a surprise. The data already existed in the United States, the United Kingdom, and Germany. What changes in France is the speed: −23.1% in nine days is a shock, not erosion.
The −0.9 slope: each point of exposure costs you one point of CTR
The study doesn’t stop at averages. It compares each domain’s CTR change to its AI Overview exposure rate. Result: a slope of −0.9. Each additional point of exposure costs roughly one point of CTR.
If 15% of your queries trigger an AI Overview, expect around 15% lower CTR. It’s rough. But it gives an order of magnitude. A slope is not a prediction. Highly exposed sites can do well. Others, lightly exposed, drop for other reasons. Your exposure rate is a risk factor, not a verdict. Ahrefs’ team says it clearly, and will return in a few months to observe real trajectories.
For e-commerce, the data reading changes. You can estimate your potential loss today by extracting your exposure rate from Site Explorer. The « AI Overview » filter in the Organic keywords report gives the exact list of affected queries.
The panel median is more reassuring: −5.7% CTR, −8.8% clicks. .fr domains lose 12.3% of clicks at the median. Local players are more exposed than international ones.
The distribution of daily click losses across the panel:
| Daily click loss | Share of panel |
|---|---|
| More than 10% | 48% |
| More than 30% | 19% |
| More than 50% | 5% |
One in two sites lost more than 10% of daily clicks. One in twenty lost half. In nine days.
Key figures
- −23.1% CTR for the most exposed domains
- −3.4% across the entire panel in 9 days
- 82.2% of highly exposed domains lost CTR
- Slope −0.9: 1 point of exposure ≈ 1 point of CTR lost
The regularity of the phenomenon is clear. The −0.9 slope is not driven by a few extreme values. It remains stable across all 963 domains. The mechanism is predictable—you can anticipate it.
For an e-commerce site with 40,000 ranked queries, a 10% exposure rate means roughly 4,000 queries where the AI Overview captures part of the click. If the slope holds, that costs 10% overall CTR. A calculation, not intuition.
E-commerce: the pages that suffer are not the ones you think
The Ahrefs study points to health as the most affected sector in France. Informational queries (symptoms, dosages, product comparisons) trigger AI Overviews at scale. Health e-commerce is on the front line: parapharmacy, nutrition, active cosmetics.
What I see among my clients is different. Product pages hold up. Buying guides, comparison pages, blog articles that answer questions do not. A product page optimized for « 600 W stand mixer » often keeps its click. A guide page « which stand mixer to choose in 2026 » loses CTR in layers.
Why? An AI Overview does not remove a page from the SERP. It answers before. The visitor already has the answer. The page remains visible, sometimes with more impressions. But in the SERP, it becomes redundant.
For e-commerce, CTR loss concentrates on informational-intent queries, not transactional ones. Queries with high volume around « how, » « why, » « which one to choose » are the most exposed. Product queries with direct purchase intent hold up better.
Content is not the issue. Architecture is. Identify pages that attract informational traffic and reorganize them. Stop producing ever more guides that will no longer receive clicks.
How many of your top 50 e-commerce queries trigger an AI Overview today?
Health is only a preview. E-commerce players in electronics, home, and fashion see the same shift on their guide content. Category pages with product lists remain protected: purchase intent is already there.
The more informational the query, the more the AI Overview is there. The more transactional it is, the more the click survives. The query portfolio needs rebalancing.
On sites with 50,000 to 200,000 pages that I deploy, buying guides lose 2 to 3 times more CTR than product pages. Order of magnitude. The global average hides the local violence.
Un cas extrême du panel français résume tout le phénomène : +183% d’impressions, mais −1,9% de clics. Voici les trois indicateurs clés ramenés à une base 100 avant le lancement.
Le découplage avant / après AI Overviews
Impressions en hausse, clics et CTR en chute — base 100 avant le 22 juillet 2026
The extreme case: +183% impressions, −1.9% clicks
The most extreme case in the French panel shows +183% impressions with −1.9% clicks. Its CTR was divided by more than three.
This decoupling is the study’s strongest signal. Rankings barely moved at the end of July. Pages continued to display, sometimes more often than before. They just received fewer clicks. The AI Overview doesn’t remove the page. It makes it useless for the query.
An AI Overview doesn’t take you out of the SERP. It makes you redundant inside it.
At one of my e-commerce clients, a product page went from 12,000 to 18,000 impressions on an informational query in ten days. Clicks stayed stable at 143. Order of magnitude. CTR plunged. Not an Analytics failure. The AI Overview captures attention above the fold.
This phenomenon has a name: the great decoupling. Impressions go up. Clicks go down. The two curves no longer intersect.
For e-commerce, this decoupling distorts dashboards. A rising impressions chart can suggest visibility is improving. But if clicks don’t follow, visibility is worthless. Track CTR by query, not just impressions.
Does your Search Console already show this decoupling on your top 20 e-commerce queries?
The Ahrefs study finds the same pattern as the global study from late 2025. Pages keep their ranking. They lose their role. A ranking without a click is a storefront on a pedestrian street: nobody stops.
For e-commerce teams, a new reflex: stop celebrating impressions. Celebrate clicks. And monitor the click/impression ratio by search intent.
Pas besoin de data scientist : la méthode tient en une poignée d’étapes dans Ahrefs. Voici le pipeline complet pour mesurer votre exposition en dix minutes.
Calculer son taux d’exposition en 6 étapes
La méthode Ahrefs exacte, sans data scientist
How to calculate your exposure in 10 minutes
No data scientist needed. Ahrefs built the exact filter. In Site Explorer, open the Organic keywords report. Select the « AI Overview » filter in SERP features. You get the list of queries where an AI Overview appears on your rankings.
Then calculate your exposure rate: number of queries with AI Overview divided by total ranked queries. Apply the −0.9 slope. A rate of 15%? Expect 15% CTR loss. A rate of 30%? The loss climbs to 27%.
The panel median sits at 5.4% exposure. And 29% of domains are already above 10%. E-commerce players that depend on guide content and comparisons are often much higher.
I run this calculation on the 50 queries that generate the most clicks, then on the 50 that generate the most impressions. These two lists rarely overlap. High-impression, low-click queries are the most exposed.
Once exposure is measured, segment: transactional queries, informational queries, brand queries. AI Overview doesn’t strike uniformly. Your action plan depends on this segmentation.
Here are the three filters I use with my e-commerce clients:
- « Which product to choose » queries: strong exposure, high CTR loss.
- « Product + reviews » queries: medium exposure, moderate loss.
- « Buy product + brand » queries: low exposure, click preserved.
This segmentation changes everything. A site that invests heavily in « best [product] » guides faces much higher exposure than a site focused on product pages and brand pages.
The calculation remains rough, but it gives an order of magnitude. An e-commerce owner wants exactly that: knowing where to focus efforts right away, before the monthly report.
Some free tools already give an approximation via GSC by filtering queries with abnormally low CTR and rising impressions. The Ahrefs filter remains the most direct.
What e-commerce players gain by acting since July 22
Before rewriting anything, I map the exposure. Sites that save time know, query by query, where the AI Overview captures the click.
Then I stop producing informational guides that no longer generate clicks. We restructure. We put budgets in the right places: product pages, transactional comparison pages, category pages built for purchase intent. No extra ad spend. Just an architecture that protects remaining CTR.
The work happens in three stages: exposure audit, internal linking overhaul, removal or merging of pages that became redundant. Semantic cocoons regain their role: concentrating authority on pages that convert.
An 8,000-page e-commerce site lost 17% CTR on 47 informational queries in twelve days. I cut production. Restructured the linking. Reinvested the next $8,000 in the right place. Result: +29% clicks on transactional pages. No ads.
This figure, I don’t sell as a promise. It’s a field observation. Every catalog has its own exposure profile. Copying a recipe is useless. What works is building a system that fits your queries.
The method hasn’t changed since 2016: semantic cocoons, not content factories. AI Overview just accelerates the shift toward pages that convert.
A ranking without a click is dead visibility. How many of your guide pages are already in that situation?
The tipping point: 5.4% today, how much in three months?
Today, the panel’s median exposure is 5.4%. 29% of domains are already above 10%. Since 2024, on every market, these numbers have moved in only one direction: up.
The impact will worsen. AI Overviews take more surface area and more confidence. Clicks will continue to concentrate on transactional queries, where the purchase is already decided.
For e-commerce, the window remains open. Sites that map their exposure now gain a three-month head start. Those who wait until the end of the year discover CTR losses already in place.
AI Overviews will touch your catalog. The question is on which pages, with what intensity, and what you do to protect the clicks that pay.
In 2026, e-commerce SEO is no longer won by content volume. It’s won on something else: spotting pages that became invisible in the SERP, restructuring them, putting authority back where the click still exists.
Today, health is dropping the most. The other e-commerce verticals will follow. Electronics, home, beauty, food: all have guides and comparisons, and each depends on informational traffic to feed the funnel.
Those who map their exposure before the September back-to-school period gain a real head start. The others will have to react urgently while the data deteriorates.
The data already exists. The tools too. Only one decision is missing: look.
What is your AI Overview exposure rate on your 20 most strategic e-commerce queries?
AI Overviews Audit: Measure Your CTR Loss in 48 Hours
I measure your real AI Overview exposure, cross-reference Search Console and Ahrefs, and show you the 20 e-commerce queries that cost you the most. I don’t sell you the method. I show you the pages.
Book a strategic call — 45 minFrequently Asked Questions
What exactly are AI Overviews and since when have they been in France?
AI Overviews are AI-written answers placed at the very top of Google results. In France, the rollout dates from July 22, 2026. Ahrefs looked at what happened on 963 domains: 28 days before, then 9 days after the launch.
How do I calculate my AI Overview exposure rate?
In Ahrefs Site Explorer, open the Organic keywords report. Filter on « AI Overview. » Divide the number of queries with AI Overview by the total ranked queries. Multiply by 100. You have your exposure rate.
Which e-commerce sectors are the most affected in France?
The study says it: health is the most affected sector. In e-commerce, buying guides and comparisons take a bigger hit than product pages. Informational queries are the most exposed across all sectors.
Is the CTR loss permanent or temporary?
Nothing is permanent. But since 2024, AI Overview exposure rates have climbed on every market. My clients who measure their exposure and rework their informational pages hold up better.
What concrete steps can you take to limit the impact of AI Overviews?
Start by looking at your exposure. Then cut informational guides that bring no clicks. Authority must go back to product pages and transactional pages. Then internal linking: it must push the pages that convert.
