According to Google, AI Overviews (AIOs) had over 1.5 billion users a month in Q1 2025. That would be 18.3% of all the people on the planet or 26.6% of all internet users. That’s a massive reach.
In traditional Ahrefs fashion, I wanted to do the largest data study ever done on AIOs to see what we could learn and share with the marketing community. For a twist on most studies, I’ve shown how they compare to search queries overall.
As a bonus, I’ll show you how you can pull the numbers yourself any time you want updated data.
Let’s dig into the data.
At the very least, AIOs aren’t likely to impact Google’s bottom line. I wonder how much they will impact companies? Companies investing in their content are also the most likely to show for AIOs. Some may even benefit from them, but the vast majority, I think, will lose traffic. We’ll have to see how it plays out.

AIOs show for 9.46% of all keywords, but they haven’t fully rolled out
At the time of writing, we have ~55.8 million AIOs from Desktop search results. This is out of 590M total keywords used to build our organic search traffic index, ~9.46% of the total. You can refer to our Big Data page at any time to get the current number of keywords in our index and the number for each country.

AIOs show for 54.61% or more of all searches by volume
You can switch the tab to Impressions to see the total search volume for all the terms showing AIOs, or their total visibility. As of writing, it’s ~98.3 billion total volume out of ~180 billion total monthly searches in our index. 54.61% of all Google searches are showing AIOs.

Sidenote.
Our data undercounts the total number of search terms and impressions for AIOs. We don’t pull SERPs for one-off or low volume queries to build our traffic index, but AIOs do show on many of these queries. We also pull AIO data from logged-out users, when they’re more likely to show to logged-in users. Many countries still don’t show them to logged-out users yet. It wouldn’t surprise me if AIOs are showing for more than 75% of all Google searches by volume.
Ahrefs and Google by the numbers
Because I don’t want anyone directly comparing our numbers to Google’s number where they say there are over 5 trillion searches annually, I feel like I need to quickly explain some differences.
Some quick math on our numbers. 180 million monthly searches * 12 months = 2.16 trillion searches a year in our data. That’s less than the 5 trillion Google says.
To try to explain the difference, there are many low-volume searches that Google classifies as anonymous queries. According to Google, 15% of searches have never been searched before. When we looked at GSC data a few years ago, 46.08% of queries were anonymous.
2.16 trillion (our yearly search numbers) / 0.54 (non-anonymous query %) = 4 trillion searches. That’s in the ballpark. Any further differences could be because anonymous queries increased, we have some gaps in our keyword database, or our volume estimates are a bit low.
Sidenote.
I do suspect the number of anonymous queries has increased and will increase further. People’s search behaviour has changed and they’re searching with much longer queries than they used to.
Google no longer has a max word count when searching. It used to be 32 words max, but now you can have as many as you want until the max URL limit is reached.

You can further filter by different data points like Intent to see more information about the type of terms where AIOs are showing.

For example, for the above search I got 24.3M keywords for Informational intent out of 24.8M total for the US. That’s 97.70% of the total.
Remember that searches can have more than 1 intent. Here’s a breakdown with the current numbers:
- Informational: 97.70%
- Navigational: 1.23%
- Commercial: 12.86%
- Transactional: 2.85%


LLMs work as next word predictors with a bit of variability built in. If you’re mentioned more in the training data such as web pages, you’re going to be mentioned more in the outputs of LLMs.
Final thoughts
Don’t wait for the next big study if you need to report what is happening with AIOs. You can run the data yourself with Ahrefs at any time with Brand Radar and Keywords Explorer.
If you have questions, message me on X or LinkedIn.
Similar Posts

Domain-Level Link Metrics May Not Be Good Predictors of AI Search Mentions
Going into this study, I suspected that domain-level link metrics would not be a good predictor for mentions. I was expecting Google AI Overviews to show more of a correlation since they may use some of the traditional search signals, whereas the other systems may not. For traditional search, usually the page-level metrics are more…

Why I Wouldn’t Start Affiliate Marketing in 2025 (And What I’d Do Instead)
Back in the day, I ran a couple of semi-successful affiliate websites. I even once ranked #1 for “best protein powder.” This was a pretty lucrative keyword given its commercial intent and high search volume. Ever since then, the idea of starting another affiliate site and getting serious about it has always been in the…

Why .store Is the Perfect Domain Extension for Your Online Shop
In the vast, ever-evolving eCommerce landscape, filled to the brim with new digital marketplaces, standing out is vital for online shop owners. And you shouldn’t underestimate the impact your top-level domain can have on your brand’s online presence. Over the last 12 months, the .store registry commissioned Contrast Digital to run a TLD experiment to…

The 7 Phases of the Search Demand Lifecycle: Going Beyond Search Volume
Many junior SEOs make the same mistake: they chase high search volume, assuming it means long-term opportunity. But search volume alone is completely misleading. A keyword with 50,000 searches/month sounds like a goldmine, until you realize it’s already in decline. Investing in a fading trend is like buying MySpace stock today. Every keyword follows a lifecycle…

What Are the Parts of a Domain Name, and What Do They Do? A Clear Guide to Knowing It All
Aristotle once said, ‘The whole is greater than the sum of its parts.” But it’s the parts of a domain name we care about. Learn about them in our guide. Even if technology isn’t your jam, you probably know how to find a particular website: you type the company name into your browser, plus a…

The Long-Form Content Illusion: Why Longer Content Isn’t Better
You’re looking for this article because you think long-form content works right? Or at least that’s what you’ve heard from many marketers. You might even have seen it for yourself. You’re trying to make a simple pasta dish (like aglio e olio) but you have to scroll through a piece that’s thousands of words long,…