AI takes clicks from organic search: how it affects traffic

How AI Overviews affect organic traffic: a direct answer

AI Overviews, or generative answers, can reduce the number of clicks from organic search results, especially for informational and non-brand queries. This is because users often get answers directly in the search results without visiting websites.

Research and statistics on the impact of AI Overviews

New studies show a significant drop in click-through rate (CTR) for organic search results with the appearance of AI Overviews. According to Ahrefs analysis, CTR for the #1 organic position dropped by 34.5% where AI Overviews appeared. Amsive recorded an average CTR decline of 15.49%, and for queries combined with featured snippets, the drop reached 37.04%.

These studies, covering hundreds of thousands of keywords, confirm that AI Overviews appear more often for non-brand, informational queries. In particular, Amsive noted a -19.98% CTR drop for non-brand queries, while Ahrefs focused on informational queries, where there is a 99.2% overlap with the appearance of AI Overviews.

In Poland, a Senuto study covering 17.7 million search queries showed that AI Overviews are already present in 24.17% of all queries. This led to a 19.4% decline in real traffic to websites in June 2025 compared with the previous year, while impressions increased by 3.3%. This is a classic example of a zero-click search, when users see more information but click less often.

AI Overviews are generated much more often for informational queries (“Know” – 31.37% of cases) than for transactional ones (“Do” – 7.16% of cases). This means that sites that traditionally attracted traffic from informational queries may feel the greatest impact.

How do AI Overviews affect rankings and visibility?

AI Overviews push organic results lower on the results page, reducing visibility even for pages that rank well. According to Amsive, for queries outside the Top 3 organic positions, CTR declined by 27.04%.

By contrast, branded queries are less likely to trigger AI Overviews (only 4.79% of cases). However, when they do appear, CTR increases by +18.68%. This may be related to stronger user intent and brand familiarity.

Do you need special optimization for AI Overviews?

Google states that no special optimization is needed for AI Overviews, since they rely on the same principles as traditional SEO: quality content, crawlability and indexability, and source authority. However, practical observations suggest the need for adaptation.

Google Search Central notes that there are no additional requirements for appearing in AI Overviews beyond those that apply to the general search index. The main technical requirements include:

• pages must be indexed;
• pages must be eligible for search snippets;
• they must meet Google’s general technical requirements.

Google also confirms that no special machine-readable files or AI-specific markup are required. More important are content quality, structure, and source authority.

Key factors for content appearing in AI Overviews

Although Google does not require specific optimization, practical observations point to certain aspects that help content get used in AI Overviews:

• Unique perspective: AI systems analyze multiple sources, so it is important to have your own distinct point of view. A firsthand review provides unique experience and context.
• Expertise and depth: Add depth and expertise rather than simply repeating existing information.
• Clarity and brevity: Information should be easy to extract and understand. Facts, lists, and tables on the page are prioritized.
• Structured content: Easy data extraction and a high density of facts.
• Source authority: Trust in the site and its content.

The introduction of AI Overviews is turning SEO into “Generative Search Optimization” (GEO). The goal of GEO is to have content used and cited in AI answers, not just ranked in a list of results.

Technical aspects and control mechanisms

For content to appear in AI Overviews, it must be crawlable and indexable. If a page is blocked in `robots.txt`, has a `noindex` meta tag, or is rendered only on the client side in a way that bots cannot access the content, it will not be considered.

Google Search Console provides a “Generative AI performance” report that shows impressions for AI Overviews and AI Mode. This report lets you distinguish a link impression from a click-through. However, data from this report should not be equated with data from other AI systems such as ChatGPT or Perplexity.

To control how content may appear in search results, Google offers the following directives:

• `nosnippet`: limits the ability to show a snippet (does not necessarily block AI Overviews, but may affect what is shown).
• `data-nosnippet`: similar to `nosnippet`.
• `max-snippet`: limits snippet length, which may affect content extraction for AI Overviews.
• `noindex`: removes the page entirely from search, including AI Overviews.

It is important to distinguish between AI Overviews and AI Mode. AI Overviews are generative overviews with links in the regular Google Search results. AI Mode is a separate search experience for more complex queries and comparisons, which may use different models and search methods.

How do you measure the impact of AI Overviews on traffic?

Measuring the impact of AI Overviews requires a comprehensive approach:

1. Analyze Google Search Console data: The “Generative AI performance” report provides impressions for AI Overviews and AI Mode. Track rising impressions alongside falling clicks, which may indicate that the AI summary is answering the question without a visit to the site.
2. Manually track queries: Make a list of the keywords your customers ask about and run them monthly through the main AI assistants. Record whether you are mentioned, cited, and which page is linked.
3. Analyze branded queries and direct traffic: Mentions without direct clicks may show up later, when users begin searching for your brand name.
4. Separate impressions from clicks: Distinguish a link impression from an actual click, lead, or sale. Compare the same pages and periods, accounting for seasonality and advertising changes.

For a deeper analysis of specific queries, it is recommended to keep a separate log recording search mode, language, country, date, AI response, and the cited URL. These observations should be combined with organic search data and business metrics.

When diagnosing a drop in organic traffic, first check technical causes (indexing issues, `robots.txt`, `noindex`, etc.) and only then blame AI summaries.