Clicks from AI Overviews: What content Google says is in demand

Liz Reid, head of Google Search, has described which content users actually click from AI Overviews. It's an important statement for any company that wins customers through content: the AI summary answers simple questions on its own, and clicks go to whatever goes further.

In this article, we put Google's statement in context, explain how Google measures click quality and show how we build content that gets clicked in AI search. Our sources are Reid's post on the Google blog and Google's July 2026 guide to AI search.

Google is shaping Search around user behavior

Google adapts Search to what users search for and click. Reid says people are increasingly visiting sites with forums, videos, podcasts and posts where they find authentic voices and first-hand experience. As a result, some sites get fewer visitors and others more.

According to Reid, total organic clicks from Google to websites have stayed roughly flat year over year. What has grown is the number of clicks after which users don't quickly return to search. These figures come from Google itself and have not been independently verified.

For your content, this means technical optimization remains a prerequisite. A page needs a recognizable reason for the click, such as its own data, practical examples or an explanation that goes beyond the AI summary.

Google judges AI content by usefulness, not origin

Google doesn't reject content because it was created with AI. Its AI search guide says such content must meet the same standards as any other. Mass-producing content to influence rankings or AI answers violates Google's policy on scaled content abuse.

So AI is allowed, but it's no substitute for substance. A text generated from a prompt and published as is usually says the same thing as countless others and gives Google no reason to surface it. We use AI for research and structure. The content readers click for comes from hands-on work with real accounts and projects.

Depth and an original perspective earn the click

The key point concerns the type of content. Google's guide recommends unique, non-interchangeable content and calls it the factor most likely to influence visibility in AI features. According to Google, interchangeable content is based on general knowledge and could have come from anyone.

Reid describes the division of labor this way: the AI answer provides an overview, and anyone who wants to learn more, go deeper or buy clicks through. Those clicks, she says, are more valuable. In concrete terms:

Depth

• Analyses with concrete figures and data
• Guides with real examples
• Case studies with results and lessons learned
• Context within the bigger picture

Perspective

• Experience from your own projects
• Insights no one else has
• Reasoned opinions, even ones that challenge conventional advice
• Industry knowledge that isn't available everywhere

Variety of formats

• Videos where a process is hard to put into words
• Charts for data
• Calculators and tools
• Interviews with practitioners

Google measures whether visitors find what they came for

Google distinguishes between clicks where users stay on the page and clicks where they quickly return to search. Reid describes a click without a quick return as a sign the user is interested in the website.

Google doesn't disclose exactly how it weights this signal. But the direction is clear: a page must deliver on what its title and description promise. So we focus on three things:

• The answer comes first: The main takeaway is visible without scrolling.
• Clear structure: Each heading tells readers what the section answers.
• Fast, mobile-friendly pages: If users have to wait or can't find their way, they go back to search.

Google favors content from people with proven experience

To get clicks in AI search, you need visible expertise. Google's guide recommends an original point of view and first-hand experience, and advises against simply repeating what others have already said.

It shows in the writing itself: “We tested this in our accounts, and here's what happened” carries more weight than “experts recommend.” Clear authorship with name, role and profile is part of it too, so readers and search engines can see who stands behind the content.

Three levers to align your content with AI search

1. We invest in depth and perspective

Google explicitly says there is no ideal length. What matters is whether content covers a topic fully and adds something others don't. So for your most important topics, we build cornerstone pages that answer every relevant question, with proprietary data, project experience and a clear, reasoned stance. We link these pages to subtopics and update them regularly.

2. We use multiple formats

According to Google, users are increasingly clicking on forums, videos, podcasts and posts with authentic voices. If you rely on text alone, you stay invisible wherever users prefer other formats. Depending on the topic, we add video, graphics, calculators or interviews and make sure your experts are visible on platforms such as YouTube and LinkedIn.

3. We make your experience visible

Experience only counts if people can see it. That means authors with names and profiles, concrete numbers instead of “significant improvements,” examples from your own work and honest accounts of what didn't work. We regularly update existing content based on new data and feedback instead of publishing it once and forgetting it.

Our article on Generative Engine Optimization and our GEO page explain how we apply this to AI search as a whole.

Conclusion: The click goes to what only you can offer

Google's message is clear: the AI summary covers interchangeable content on its own. Clicks go to those who offer depth, an original perspective and real experience.

For your content strategy, this means fewer new pages and more substance in the ones you have. AI is a tool; the content comes from your work. And that content is exactly what competitors can't replicate with a prompt.