Skip to content
SEO Madmanby Adam Hafez

Google AI Mode adoption and zero-click rate, Semrush 2025

Semrush clickstream data on Google AI Mode's first two months: usage grew about 4x to just over 1% of sessions, and 92-94% ended without an external click.

Published · 3 min read

Written byAdam HafezUpdated
Share
Markdown
Close-up of hands typing a query on a laptop keyboard at a table

Key takeaways

  • Semrush found AI Mode's share of Google search sessions grew roughly 4x, from 0.25% in early May 2025 to a little over 1% by early July 2025.
  • On average only 6-8% of AI Mode sessions led to a visit to an external domain, so 92-94% were zero-click.
  • Semrush cites separate earlier studies putting zero-click at about 34% without an AI Overview and about 43% with one, but warns the comparison is imperfect.
  • The average AI Mode query ran 7.22 words against 4.0 words for traditional Google search, and sessions took 2-3 searches.
  • The data covers US desktop sessions only, from May 1 to July 5, 2025, so it says nothing about mobile or later months.

Google AI Mode is the conversational answer layer Google added to Search in 2025, and the first question publishers asked was how much traffic it sends out. Semrush published the earliest large clickstream read on that in July 2025, and it remains a primary source for how early users behaved.

How big was Google AI Mode in its first two months?

Small, but growing quickly. Semrush reports that AI Mode’s share of Google search sessions increased roughly 4x in the two months after the May 2025 rollout, from 0.25% in early May to a little over 1% by early July. Google began rolling AI Mode out in the US on May 20, 2025 and launched it for all US searchers on June 27, 2025, per the same report.

How many AI Mode sessions sent traffic to other sites?

Very few. On average 6-8% of AI Mode sessions led to a visit to an external domain, so 92-94% were zero-click. Semrush sets that beside about 34% zero-click for Google Search without an AI Overview and about 43% with one. It flags that those two figures come from separate studies with different methods and timeframes, so the gap is a trend signal, not an exact multiple.

How did AI Mode queries differ from regular searches?

Longer and more resolved. The average AI Mode query was 7.22 words against 4.0 for traditional Google search, and Semrush observed that AI Mode users had fewer daily sessions and fewer searches per session, with sessions taking 2-3 searches. Those rates stayed stable through the period. In the same comparison Semrush puts ChatGPT prompts at about 23 words, so AI Mode sits between keyword search and a chatbot conversation. Semrush reads the shift as users writing full-sentence questions rather than keyword commands.

Google’s documentation describes AI Mode as most helpful for queries that need “further exploration, reasoning, or complex comparisons.” Both AI Mode and AI Overviews may use a “query fan-out” technique, issuing multiple related searches across subtopics and data sources, and Google says this lets them show a wider and more diverse set of links than a classic search. It also notes that the two features may use different models and techniques, so the responses and links they show will vary. Google says there are no extra requirements or special markup for either: a page must be indexed and eligible to show with a snippet.

How can a site measure its own AI Mode traffic?

Not separately in Search Console. Google says sites appearing in AI Mode are included in overall search traffic in the Performance report, within the “Web” search type. It suggests tracking conversions and time on site in other tools, such as Google Analytics, to judge the value of those visits. Semrush’s own advice leans the other way, toward visibility over clicks: build topical authority across several angles, and watch brand mentions and sentiment inside AI answers. For how click-through changed on classic results with AI Overviews, see our study on AI Overviews and position-1 CTR, and for the wider picture, zero-click Google searches.

What are the limits of this data?

The sample is US desktop sessions from May 1 to July 5, 2025. It does not cover mobile, other countries or anything after early July 2025, so it cannot show how AI Mode traffic behaves now. A rising usage share and a low external click rate together describe early adopters, not the whole market.

The evidence

Period
May 1 to July 5, 2025
Sample
Almost 69M Google Search sessions, US desktop

Method: Semrush analysed anonymised clickstream data for almost 69M Google Search sessions between May 1 and July 5, 2025, restricted to sessions in the US on desktop devices. It measured daily AI Mode searches and sessions, prompt length and traffic to external domains. Semrush notes that its zero-click comparisons to regular search and AI Overviews come from separate studies with different methodologies and timeframes, so they should be compared with caution.

Zero-click share of sessions, as reported by Semrush
Google Search, no AI Overview~34%
Google Search, with AI Overview~43%
Google AI Mode~93%

Sources

  1. 1.Google AI Mode's Early Adoption and SEO Impact - Semrush, July 30, 2025Primary
  2. 2.AI features and your website - Google Search CentralPrimary

Frequently asked questions

What share of AI Mode sessions send users to another website?

Semrush found that on average 6-8% of AI Mode sessions led to someone visiting an external domain, which means 92-94% of AI Mode searches were zero-click.

How fast did Google AI Mode grow in its first two months?

In Semrush's sample, AI Mode's share of Google search sessions rose roughly 4x, from 0.25% in early May 2025 to a little over 1% by early July 2025.

Can the AI Mode zero-click figure be compared directly with regular search?

Not cleanly. Semrush says the regular-search and AI Overview figures come from separate studies with different methodologies and timeframes, and asks readers to compare with caution.

About the author

Adam Hafez
Adam Hafez

Founder, UpgradIQ FZC LLC

Adam Hafez works on technical SEO and search measurement: how pages get crawled, indexed, ranked and now quoted by answer engines. He founded UpgradIQ, which reads Google Search Console and GA4 to tie ranking movement back to the changes that caused it. He publishes what the data supports and states the limits of it.

  • Technical SEO
  • Search Console and GA4 measurement
  • Answer engine optimization
  • Structured data

The briefing

Only what actually changed in search, delivered in full by RSS, Atom or JSON feed.

Follow