Natural referencing remains the main traffic acquisition lever for a website. However, the way this traffic is generated has changed: since the deployment of AI Overviews by Google in France on July 22, 2026, being well-positioned in the results no longer produces the same volume of clicks as before. What indicators should be monitored to measure a site’s actual visibility, and what levers should be acted upon when the organic click-through rate declines?
Organic Click-Through Rate Before and After AI Overviews
The data published since 2024 allows us to quantify the impact of AI-generated summaries on organic traffic. The table below summarizes the results of three independent studies.
| Source | Scope | Measured Impact |
|---|---|---|
| Pew Research Center (July 2025) | 68,000+ queries (United States) | Organic click-through rate: 8% with AI summary versus 15% without |
| Ahrefs (April 2025) | 300,000+ informational keywords | Approximately 35% drop in click-through rate in first position with AI Overviews |
| Define Media Group (May 2024) | 64 news sites | Decrease of 42% in traffic on affected queries |
These discrepancies show that visibility in search results and actual traffic to a site are now two distinct metrics. A site can occupy the top position on Google and receive half as many visitors as it did two years ago for the same query.
For companies that rely on natural referencing, the available data on https://www.referencement-consulting.fr/ provides additional insights into adaptation strategies in this context.
Informational Queries and Transactional Queries: A Growing SEO Profitability Gap
AI Overviews do not affect all queries in the same way. The Ahrefs study specifically focuses on informational keywords, where the user seeks a factual answer (“what is SEO,” “how to improve a site’s speed”). It is on this type of query that the summary generated by Google directly responds, without the user needing to click.
Transactional or navigational queries (searching for a provider, a product, a local service) maintain a significantly higher click-through rate. Google cannot “summarize” a purchase or a contact.
The consequence for a content strategy is direct:
- Pure informational pages (guides, definitions, FAQs) lose organic traffic even if their positioning remains stable. Their role evolves towards building thematic authority rather than directly generating visits.
- Pages targeting an action intent (requesting quotes, comparing services, local searches) maintain their yield in terms of clicks and conversions.
- The internal linking between informational content and transactional pages becomes the mechanism that transforms authority into useful traffic.
Technical Signals Affecting Natural Referencing in 2026
Google’s algorithm incorporates technical criteria whose relative weight has evolved. Two areas deserve particular attention because they condition both indexing and eligibility for AI Overviews.
Structured Data and Eligibility for AI Summaries
Sites whose pages use structured markup (schema.org) appear more frequently as cited sources in AI Overviews. Being cited in a generated summary does not produce the same click as a traditional blue link, but it is now the main form of visibility for informational queries.
FAQ, HowTo, and Article markup remains relevant. However, structured data does not compensate for weak content: it facilitates information extraction by the engine, not value creation.
Web Performance and Core Web Vitals
Loading speed, visual stability, and page responsiveness remain ranking criteria. A slow site loses positions to a fast competitor with equivalent content. The measurement tools provided by Google (PageSpeed Insights, Search Console) help identify pages that slow down the entire site.
SEO Content Facing Generative Engines: What Still Works
The temptation is to produce more content to cover more keywords. Recent data suggests the opposite: the depth of content matters more than the volume of published pages.
Google penalizes mass-generated AI content when it does not provide added value compared to what already exists. The E-E-A-T criterion (Experience, Expertise, Authoritativeness, Trustworthiness) is applied more rigorously since the updates of 2024-2025.
Three characteristics distinguish content that retains its traffic despite AI Overviews:
- They address a specific intent with original data (experiential feedback, measurements, concrete cases) that the engine cannot synthesize from other sources.
- They target mixed-intent queries, where the user needs to compare or evaluate before acting, not simply obtain a definition.
- They are structured for the web (clear headings, short paragraphs, semantic markup) while offering an editorial angle distinct from competitors.
An article that rephrases what ten other sites already say no longer has a reason to be clicked when Google displays the summary at the top of the page. Natural referencing now rewards information that the engine cannot replace.
Organic traffic remains the acquisition channel least dependent on a recurring advertising budget. Sites that adjust their content strategy to the new conditions of Google search, targeting queries where the click still holds value, maintain a measurable advantage over those that continue to optimize as they did in 2020.