AI Overviews cut CTR by 23.1% in France. Here's what that data means for SEO teams in Southeast Asia — and how to adapt before the same wave hits.
France got Google’s AI Overviews on July 22, 2026. Within nine days, the most exposed domains had already lost 23.1% of their click-through rate. That’s not a projection — that’s Ahrefs tracking 963 domains in Google Search Console, before and after.
For SEO leads across Southeast Asia watching this from a distance: don’t. This is the preview screening.
The CTR Collapse Is Structural, Not a Blip
Ahrefs’ France study is the clearest controlled data we’ve had on AI Overview impact. The methodology matters — 28 days pre-launch versus 9 days post, same domains, same query sets. The 23.1% CTR decline among the most AI Overview-exposed domains isn’t noise; it’s the model doing exactly what it was designed to do: resolve queries without a click.
The strategic implication is uncomfortable but worth saying plainly: a significant slice of informational search traffic is being reclassified from “yours to earn” to “Google’s to answer.” This isn’t a penalty. It’s a structural redistribution. Brands that built authority on how-to content and definitional queries are disproportionately exposed. In Southeast Asia, where mobile-first users on Grab, Shopee, and LINE often begin product research with broad informational queries, the blast radius could be larger than in Western markets once AI Overviews roll out regionally.
Being Cited Is the New Ranking
Here’s the flip side that most post-mortems on AI Overview impact bury in paragraph seven: the domains cited inside AI Overviews don’t lose traffic the same way. In some query categories, citation exposure drives brand recall and downstream navigational search — which means entity authority is now a first-order SEO asset, not a secondary consideration.
What determines citation eligibility? Structured semantics, E-E-A-T signals, and — increasingly — how clearly your content defines its own entity relationships. Google’s crawler guidance updates, including recent expansions to Mediapartners-Google documentation reported by Search Engine Journal, signal that Google is continuing to build richer content classification infrastructure. The implication: how your content is understood by crawlers is becoming as consequential as how it ranks.
Practically, this means audit your top-traffic informational pages for schema markup completeness — FAQ, HowTo, Article, and Speakable schemas are not optional decoration anymore. They’re the structured signal that tells a generative engine what your content is about, not just what it contains.
AEO and GEO Aren’t Replacements — They’re the New Floor
Answer Engine Optimisation and Generative Engine Optimisation have been discussed as future-state strategies for the better part of two years. France 2026 is the moment they graduate from strategic horizon to operational requirement.
The practical distinction: AEO focuses on how your content gets structured and surfaced in direct-answer formats (featured snippets, AI Overviews, voice responses). GEO extends that logic to generative AI interfaces — ChatGPT, Perplexity, Gemini — where citation and brand mention frequency shape how confidently a model references your brand without a prompt to do so.
For Southeast Asian brands, there’s a compounding consideration: multilingual content architecture. A brand operating in Thai, Bahasa Indonesia, and Vietnamese needs entity authority established per language and per market, not just in English. LLMs trained on lower-resource languages have thinner entity graphs — which means brands that invest in structured, authoritative content in local languages now are building a moat that will be genuinely hard to close later.
One concrete starting point: map your top 20 informational pages against the queries most likely to trigger AI Overviews in your category. For each, identify whether you are the answer, a source for the answer, or neither. That triage drives your next 90 days of content and schema work.
What Google’s Crawler Guidance Shift Actually Signals
The recent update to Google’s Mediapartners-Google crawler documentation — expanding its scope beyond AdSense-specific guidance, as reported by Search Engine Journal’s Roger Montti — is easy to file under “technical housekeeping.” It’s worth reading more carefully than that.
When Google broadens the framing of a crawler that has historically been about monetisation signals, it suggests the underlying classification systems are being retooled for a content ecosystem where the relationship between content, crawling, and revenue pathways is more complex. Programmatic content, AI-generated pages, and hybrid editorial workflows are straining old taxonomies.
For SEO teams, the practical read: Google is actively updating how it thinks about content classification infrastructure. Teams still relying on crawl configurations and robots.txt logic built two or three years ago should schedule a technical audit. Not because something is broken — but because the system they configured for has quietly changed around them.
The brands that will hold ground in AI-mediated search aren’t the ones who optimised hardest for the old ranking model. They’re the ones who understood early that discoverability is now an entity game — and started building entity authority before the traffic graphs made it obvious.
When AI Overviews arrive in your market, will your brand be the answer, or just another result that didn’t make the cut?
At grzzly, we work with marketing and growth teams across Southeast Asia to build search strategies that account for where discoverability is actually heading — not where it was two years ago. If your team is navigating the shift from traditional SEO to AEO and GEO, we’d like to think through it with you. Let’s talk
Sources
Written by
Sneaky GrizzlyTracking the quiet revolution inside LLM-powered search — where brand mentions, structured semantics, and entity authority rewrite the rules of discoverability before most marketers notice.