Google AI Overviews

Q: Do AI Overviews harm my website traffic?

They can cost clicks, because some users already find their answer in the summary. At the same time they bring visibility and trust when you appear as a named source. The goal is therefore to be cited in the Overviews rather than to ignore them.

Q: How do I increase my chance of being named in the AI Overview?

Answer specific questions directly and concisely, use clear headings and definitions, back up statements with current facts and assign content unambiguously to your brand. Structured data and high professional credibility help the AI classify you as a reliable source.

Google AI Overviews are the AI-generated answer boxes Google places above the classic blue links on a results page. A Gemini-family model reads a handful of retrieved pages, writes a short summary, and shows cards linking to the sources it drew on. As of mid-2025 they reached over 2 billion monthly users across 200+ countries and 40 languages, which makes them one of the largest single surfaces in search — and one where showing up as a cited source now matters as much as ranking does.

Why it matters for your visibility

AI Overviews sit between the searcher and your site, and that changes the economics of a click. Pew Research Center, working from real browsing data across roughly 69,000 Google searches, found people clicked through to a traditional result in only 8% of visits when an AI summary appeared, versus 15% without one — about half the click rate. Ahrefs, analyzing 300,000 keywords over two years of Search Console data, found AI Overviews correlate with a 58% lower average click-through rate for the #1-ranking organic result. Zero-click behavior overall has also risen: SparkToro and Similarweb put the share of US Google searches ending with no click to any website at 68.01% in early 2026, up from 60.45% in 2024. None of that means traffic is dead. It means a page that never gets named in the summary loses attention it used to get just by ranking well, while a page cited inside the Overview gets seen even without a click.

How AI Overviews work

Technically, Google's Gemini-family models sit behind AI Overviews. When you enter a query, Google first retrieves a set of matching web pages, then a language model condenses their core claims into a new, coherent answer rather than inventing one from memory — the retrieval-augmented generation pattern. Alongside the text, Google shows small cards linking to the pages it used. Overviews trigger most often for queries that need explanation: comparisons, how-tos, definitions, troubleshooting. For simple navigational or brand-name searches they appear less often, because there is nothing to summarize.

Common mistakes

The biggest misconception is that a top ranking guarantees a mention in the Overview. It does not: Google's own guidance on optimizing for its AI features states plainly that no special markup, schema, or AI-specific files are required, and explicitly warns against writing separate content "for AI." A related mistake is publishing an llms.txt file expecting it to influence inclusion — Google's John Mueller confirmed in 2025 that no Google Search system reads or acts on llms.txt, and an Ahrefs analysis of roughly 137,000 sites that published one found about 97% saw zero measurable referral traffic tied to it. What actually holds pages back is plainer: vague marketing copy without concrete facts, no clear headings or short answer paragraphs for the model to extract from, and stale figures or prices that a model summarizing current information is less likely to reuse.

Relation to AI recommendations

AI Overviews are one visible instance of a broader shift researchers call generative engine optimization, a term introduced in the 2024 ACM SIGKDD paper "GEO: Generative Engine Optimization" by Aggarwal, Murahari, Narasimhan, Deshpande, Rajpurohit and Kalyan. But the systems behind it don't all select sources the same way. Ahrefs found only 6-8% of URLs cited by ChatGPT overlap with Google's top-10 for the same query, and about 80% of ChatGPT-cited URLs don't rank in Google's top 100 at all — AI citation is a genuinely different selection process from search ranking, not a rebadged version of it. One lever does seem to travel across engines: Ahrefs found third-party brand mentions correlate with AI citation rate at roughly 0.664, about three times stronger than backlinks at 0.218, which suggests getting talked about elsewhere on the web matters more here than link-building alone.

Example

Imagine you run a bike shop and someone searches "storing an e-bike battery correctly over winter." In the past, a list of guide links appeared. Today an AI Overview sits at the top summarizing the key steps — cool, dry, roughly 50 percent charge — with a handful of source cards beneath it. If your site has a clear, well-structured page covering exactly that, your shop can end up as one of those sources. The searcher sees your name, associates it with the answer, and may come to you for the next purchase without ever clicking through today.

Common questions

Do AI Overviews harm my website traffic?

They can reduce clicks — Ahrefs measured a 58% lower CTR for #1-ranking pages when an Overview appears, and Pew found click-through roughly halves. But being named as a source still delivers visibility and trust even without a click. The goal is to get cited in the Overview, not to ignore it.

How do I increase my chance of being named in the AI Overview?

Answer specific questions directly, use clear headings and short definitional paragraphs, back claims with current facts, and make sure content is clearly attributable to your brand. Google says no special schema or AI-specific markup is required, and an llms.txt file won't help — Google Search doesn't read it. Being mentioned elsewhere on the web correlates with citation more strongly than backlinks do, so digital PR and genuine third-party coverage matter.