Do AI Overviews Reduce Clicks? What the Data Shows and How to Check

What does independent data show about AI Overviews and clicks?

The clearest independent data on AI Overviews and clicks comes from Pew Research Center, which tracked the real browsing of 900 US adults in March 2025. People clicked a traditional result less often when an AI summary appeared, and they rarely clicked the summary’s own links.

Pew’s short read of 22 July 2025 is worth reading in full. The panel members agreed to share their browsing, so the data shows behaviour, not survey answers.

The main figures:

• With an AI summary on the page, users clicked a traditional result in 8% of visits.
• Without a summary, that rose to 15% of visits, nearly double.
• A link inside the summary itself was clicked in 1% of visits.
• Users ended their browsing session after 26% of pages with a summary, against 16% of pages without one.

Two limits matter. The panel was American, and the data is from March 2025. Indian search behaviour and later versions of AI Overviews may differ.

Google tells a different story. Its AI features and your website guide says clicks from results pages with AI Overviews are “higher quality”, meaning people spend more time on the site.

Both can be true at once. Fewer clicks, with the remaining clicks from people who already know what they want, is a common pattern Digital Magma sees in client reports.

Which searches lose the most clicks to AI Overviews?

Searches that ask a simple factual question lose the most clicks to AI Overviews, because the summary can settle them without a visit. Definitions, quick how-to questions and “what is” searches are most exposed. Searches for a supplier, a price quote or a booking lose far less.

Pew’s data supports the length pattern. Only 8% of one- or two-word searches produced an AI summary. For searches of 10 words or more, the figure was 53%.

Searches phrased as questions or full sentences also produced summaries more often. Those are exactly the long informational searches many content strategies were built on.

Overall, Pew counted an AI summary on 18% of the Google searches in its study. So most searches still showed none, at least in March 2025.

A rough guide, by intent:

The table is our reading of client data and Pew’s findings, not a measured study. Test it against your own queries.

Digital Magma treats the high-risk rows differently from the rest. Those AI Overviews searches still build awareness, so we keep the pages, tighten the answer and add a reason to visit. We stop measuring them mainly by clicks.

Which sites can gain from AI Overviews?

Sites can gain from AI Overviews when they are cited as a source, when their page adds what the summary cannot, or when the buyer still needs a supplier. A cited link puts a page in front of searchers who might never have scrolled to its old ranking position.

Google makes this case itself. Its May 2024 launch post said links included in AI Overviews get more clicks than if the page had appeared as a traditional listing for that query.

That is a statement about cited pages, not about all pages. It fits with Pew’s finding that overall clicks fall.

Three kinds of sites tend to do well:

• Sites with original data or experience. Google’s guide to generative AI features asks for “non-commodity” content, such as a first-hand review rather than a summary of others.
• Local service businesses. A summary can explain a treatment, but the reader still has to book a clinic.
• Manufacturers and B2B suppliers. A buyer can learn the basics from a summary, then still needs a quote, a spec sheet and a sales contact.

Digital Magma has seen the same thing with industrial clients. A buyer who reads an AI Overviews summary about welding processes still has to find a manufacturer and ask for a quote.

Our AI Overviews entry explains how the summaries choose and show their links.

How do you check your own AI Overviews click change in Search Console?

To check your own AI Overviews click change, compare clicks, impressions and CTR in Search Console for matched date ranges, query by query. Then open the Generative AI performance report to see which pages earn impressions in AI Overviews and AI Mode. Three steps take under an hour.

1. Compare dates. In the Performance report, compare the last three months with the same three months a year earlier. That removes most seasonal swings.
2. Sort by query. Look for queries where impressions held or rose while CTR fell. That pattern points to a feature, often an AI Overview, taking the click.
3. Open the Generative AI report. Google’s Generative AI performance report shows impressions from AI Overviews and AI Mode, by page, country, device and date.

Two counting rules help you read the numbers. Google’s Search Console help on performance data says an AI Overview takes one position, and every link in it shares that position.

A click on a link inside the overview counts as a normal click. So AI Overview traffic sits inside your Web search totals, not in a separate bucket.

Google says the report needs enough impressions to appear. Smaller sites may see nothing there yet.

When Digital Magma runs this check for a client, the AI Overviews pattern is usually narrow. A handful of definition and how-to pages carry most of the lost clicks. Service and location pages often look flat or up.

That matters for the decision that follows. A site-wide panic about AI Overviews is rarely justified. A focused rewrite of ten or twenty pages often is.

Keep a simple log as you go. Note the query, the date range, the click change, the CTR change and whether you saw an AI Overview when you searched it yourself.

What should you change on pages that lost clicks to AI Overviews?

On pages that lost clicks to AI Overviews, add what a summary cannot give: first-hand detail, current prices, original examples, tools and a clear next step. Then make the core answer easy to cite, so the page has a chance to appear as a source inside the overview.

Start with the pages where impressions stayed but CTR dropped. Those still rank, so they are closest to earning a citation.

Digital Magma works through each one with five questions:

1. Does the first paragraph answer the query? A direct answer near the top is easier for any system to cite.
2. What do we know that others do not? Add a result, a photo, a worked example or a price range you have actually charged.
3. Is the next step obvious? A calculator, a booking link or a quote form gives a reason to click.
4. Is the text in the HTML? Google says important content should be in textual form.
5. Is the page fresh? Update dates, figures and screenshots where they changed.

Avoid one tempting move. Google’s guide warns that making a page for every fan-out query, mainly to manipulate AI answers, breaks its scaled content abuse policy.

Our query fan-out entry explains what those related searches are.

Can you opt out of AI Overviews, and should you?

You can opt out of AI Overviews and AI Mode with the Search generative AI control in Search Console, rolled out worldwide on 31 August 2026. Most businesses should not, because excluding the site also removes every chance of being cited, while summaries still draw on other sites.

Google’s Search generative AI control help page sets out what exclusion does:

• Links to your site and your content stop appearing in Search generative AI features.
• Content from other sites still appears there, and may look similar to yours.
• Google says the control is not a ranking signal for the rest of Search.
• It does not affect AI training. Google points to Google-Extended for that.

Google says changes usually take effect within one to two days, though caching can delay some content.

Finer controls also exist. The nosnippet, data-nosnippet and max-snippet rules limit what Google can show from a page. Noindex removes it from Search entirely.

The trade-off is plain. Opting out stops the summary quoting you. It does not bring the searcher back to the blue links.

What Digital Magma has seen with clicks and AI Overviews

Digital Magma’s clients that kept growing clicks through the AI Overviews period share two traits: strong rankings on buyer searches and pages with facts only they could provide. Informational traffic softened in places, but the searches that lead to enquiries held up.

A Toronto ENT clinic, a client for four years, earned about 20K clicks in one three-month period. Patients still need to book a real appointment, whatever the summary says about symptoms.

A Canadian welding manufacturer, with us for three years, holds 40+ keywords in Google’s top 10. It is visible in Google AI Overviews and ChatGPT.

Over three months, that site earned about 30K clicks from about 100K impressions. Being cited in the overview and ranking below it, for the same buyer queries, worked together.

We do not claim AI Overviews caused those numbers. They show that a site with deep, specific pages can keep earning clicks while summaries appear above it.

AI Overviews and leads: measure enquiries as well as clicks

AI Overviews change the meaning of a click, so judge results by enquiries, calls and sales as well as sessions. A page that loses casual readers but keeps its buyers may be doing better for the business, even when the traffic graph looks worse.

Set up three numbers in GA4 or your CRM, and review them monthly:

• Leads per organic session. If clicks fall and this ratio rises, the lost visits were mostly readers, not buyers.
• Leads by landing page. This shows which pages still produce business after AI Overviews arrived.
• Branded searches. More people searching your name suggests summaries and citations are building awareness.

Digital Magma reports these beside clicks for every client. A drop in clicks gets a page-level explanation, not a shrug.

Manish Tiwari described the outcome that matters in a Google review: “Within a few months, our traffic and lead quality improved noticeably.”

AI Overviews and clicks: frequently asked questions

These are the AI Overviews questions site owners ask Digital Magma when Search Console shows impressions up and clicks down. Answers rely on Pew’s measured data and Google’s own documentation, and say which is which. Where the two disagree, both views are given.

Related AI Overviews questions, sources and author

These pages go further on AI Overviews, measurement and the content that earns citations. Digital Magma links live knowledge-base terms directly and marks answers that are still planned. Read the AI Overviews entry first, then the measurement answer once it is published, since the two work as a pair.

• Google AI Overviews: how the summaries work and pick links
• AI citation: what counts as a source link in an AI answer
• Query fan-out: the related searches behind each summary
• Google AI Mode and Google-Extended (planned entries)
• Does SEO still matter in AI search? and How to measure AI visibility

Sources: Pew Research Center’s short read of 22 July 2025, Google Search Central’s AI features guide and generative AI optimisation guide, Google’s May 2024 AI Overviews launch post, and Search Console and Google Ads help pages. All were checked on 22 October 2026.