AI Overviews and CTR in 2026: What Happens to Your Clicks

Bpract TeamUpdated on: August 10th, 2026
AI Overviews and CTR in 2026 — what happens to your clicks

Two years ago we asked how Google's Search Generative Experience would change click-through rates. The experiment has since shipped to everyone as AI Overviews, and the question is no longer hypothetical: the AI answer now sits above the traditional results for a large share of informational searches. This follow-up to our SGE and CTR article looks at what actually happened to clicks — and what to do about it in 2026.

How do AI Overviews affect CTR?

AI Overviews reduce organic click-through rates because Google answers the query directly at the top of the results page. Industry studies through 2025–26 report organic CTR falling by roughly a third to two-thirds on informational queries that trigger an AI Overview, while pages cited inside the answer keep earning clicks and brand visibility.

From SGE to AI Overviews

Search Generative Experience was Google's opt-in lab experiment. In 2024 it became AI Overviews and rolled out by default; by 2025 it covered a large share of informational queries, joined by an end-to-end conversational "AI Mode". The layered result page we described in the SGE article is now simply how search looks: an AI-generated summary with citations on top, ads and classic blue links below.

What the data shows

Across the studies published through 2025 and 2026, three patterns repeat:

  • Fewer clicks overall. When an AI Overview appears, organic results below it lose a substantial share of their clicks — commonly reported between a third and two-thirds, worst for purely informational queries.
  • Citations concentrate value. The links quoted inside the AI answer still get visits, and they reach users with genuine intent. Being source material now matters more than ranking #3.
  • Commercial and local queries hold up better. Searches with buying intent, local intent ("software company in Calicut"), or brand names still produce clicks, because the user needs a business, not a paragraph.

Meanwhile a second channel has appeared: assistants like ChatGPT, Perplexity, and Gemini send small but fast-growing referral traffic of their own — on this site, those sessions grew about 40% month over month this summer.

How do you get cited in AI Overviews?

Publish answer-first content: a direct, self-contained answer high on the page, mirrored in FAQ structured data, with clear headings and named entities. Keep pages crawlable for AI bots, maintain an llms.txt file, and build topical authority — the same fundamentals that win featured snippets win AI citations.

What this means for your content strategy

  • Answer first, elaborate second. Every important page should answer its core question in one tight paragraph near the top, backed by FAQ structured data.
  • Target queries that still click. Comparison pages, local pages, tools, and product content keep their CTR; pure definitions increasingly do not.
  • Measure citations, not just rank. Track AI-assistant referrals and whether your pages are quoted, alongside classic position reports.
  • Stay crawlable to AI. Allow reputable AI crawlers and publish an llms.txt so engines can find your canonical answers.

Conclusion

SGE did what we expected: it kept the traffic it could answer and passed on the traffic it could not. In 2026 the winning strategy is not fighting the AI answer but becoming its source — answer-first pages, structured data, and content aimed at the searches where a human still needs a business. That is how we approach our own pages, and it is working: this site's AI-referral traffic and top-5 rankings both grew while overall impressions fell.