By Last quarter, we published an extraordinary analysis of hundreds of millions of Google searches alongside our clickstream panel partners at Datos (A Semrush Company). That analysis opened many marketers and creators’ eyes to the reality of how people use Google in 2024, i.e. navigational-heavy, and significantly weighted toward the top few thousand queries.
This quarter, we’re back to examine several key questions that arose from both that research, and (next week) the ongoing fear (hope?) that Google search is losing ground to AI tools. Thanks, once again, to the incredible generosity of the Datos team, we’re able to present credible, high-quality numbers that solve long-held mysteries of how people engage with Google.
Briefly, before we do that, an important caveat: This study and Datos’ panel measure web browser activity. That means searches performed on platforms like the Google Maps mobile app, or those done in Google’s search app for Android aren’t included. Thus, for this study, we’ve concentrated on desktop activity–growth rates on desktop closely parallel what happens on the observable mobile web, but can’t take into account what’s going on in mobile apps. Nevertheless, it’s an excellent baseline to measure change over time, and it’s certainly indicative of desktop search behavior and any loss to AI tools, which happens primarily on desktops and in web browsers.
How many searches do Americans perform on Google?
There’s a lot of bad data on the Internet taken from estimates that cite questionable sources endlessly in a circle, but thanks to Datos’ team, we can finally provide concrete numbers.
Active American desktop web users in 2024 performed, on average, 126 unique Google searches each month. The median was 53 Google searches/month. If we use three other facts we know: that Google has a 90.15% market share in search globally (Statcounter), that 5.35 billion humans are active on the Internet each month (GWI), and that ~81% of Internet users perform use a search engine at least once a month (GWI), and one assumption (that mobile and desktop volumes are similar) we can estimate that Google serves at least 492 billion searches per month (or ~1.3B per day).
In a move that I can only consider serendipity from the data gods, Google, a few hours prior to the publication of this post, announced that they had “more than 5 Trillion searches in 2024.” Our math above puts the number at 5.9 Trillion, a little high, likely because Datos’ panel focuses on wealthier countries where more search activity per person is to be expected. Still incredible that they’d come out with numbers the day we publish that help back up the veracity of these results, and the quality of Datos’ panel.
Let’s break down our estimates here for clarity:
- Active American web users refers to only folks in Datos’ panel who had search activity during the months we looked at. That means Datos excluded devices that didn’t log onto the web or didn’t search Google even once.
- A “unique” Google search is different than “any and all” Google searches – Datos excluded counts of searches that were for exactly the same term on the same day. If you searched for “Tacos Al Pastor recipe” 6X last Tuesday (like I did), that would only be treated as a single “unique search” for the purposes of this study.
- Google searches include queries that started on Google.com, and those that used the “big 6” tabs in Google’s search interface: Images, Video, Maps, News, Shopping, and the newcomer, Web. We didn’t include things like Google Scholar, Forums, Flights, or other rarely/conditionally-triggered vertical options.
126 unique desktop searches/month on average may seem high to some, and low to others (pretty sure I’ve done more searches than that over some busy weekends 😅). But, the average doesn’t tell the whole story. Thankfully, this dataset lets us visualize the breakdown of search usage:
The data illuminates a fascinating reality about Google search: about 1/3rd of active web users don’t use Google all that much (only 1-20X searches/month), another 1/3rd are moderately active (with 21-100 searches/month), and a final third are very heavy searchers (performing 101-1,000+ searches each month). I checked my own search behavior and, at least for the last few months, fell into the 500-1,000 searches/month bucket. My guess is many of us who live, eat, sleep, and breathe digital marketing and the web for our jobs fall into those upper tiers, and we might sometimes forget about the folks who turn to Google only a few times each month.
Which Google search verticals (Images, Maps, Video, Shopping, etc.) are most popular?
My next line of inquiry was into the major search verticals: News, Images, Maps, etc. Many of y’all who read the original research piece asked about this, and the Datos team was incredibly kind to go through the effort of providing this information.
It’s fascinating to see the breakdown:
Remember that Maps is almost certainly undercounted here due to the mobile app issues. Also, remember that many Maps searches happen inside Google.com’s default interface. Google even shows Maps results inside the standard web interface, so there’s almost certainly a ton of “Maps” and “Local” search activity taking place; it just doesn’t start specifically on Maps.Google.com or google.com/maps in a web browser (which is what we’re counting here).
Google Images is a big story I’ve talked about for years (recall that in my old 2018 study with Jumpshot, I noted that Google Images was bigger than even YouTube in search volume!). Back then, Images had almost 1/3rd the volume of Google.com, though this was desktop-only data. In 2024, it’s about 1/8th the size of Google.com, which is still extraordinary.
If you’re a search or content marketer and you’re not using visuals (AI-generated ones don’t count 🙄) in your content strategy, now’s your wake-up call to rethink that.
And if you’re thinking those other numbers (Shopping, News, Video, etc.) look awfully small, remember that content from each of those tabs also shows in the main Google search results. This breakdown is looking at the searches that happen in those tabs/sections, not the ones that simply result in a click on a Google News or Shopping result that appeared in the default Google search tab.
How do the average and median search numbers on Google compare?
The overall average of ~126 Google searches/searcher/month (across all verticals) is dramatically larger than the median of 53. There are a lot of people searching Google hundreds of times each month and many more searching only a handful, hence this wide gap.
When we broaden that to look at all of the search verticals, a similar pattern emerges. Google.com (the default home page) gets the lion’s share with 110 searches/searcher/month, but the median is less than half. I’ve visualized all of the individual search tabs below:
Google’s new “Web” tab (which launched in May of last year and was championed by Search Liaison Danny Sullivan) has the most dramatic spread: a few people love that feature, and almost everyone else ignores it entirely. The other tabs aren’t dissimilar, and my guess is that a large number of Google’s searchers rarely, if ever, click to another tab/vertical to perform their search.
Side note: If you’re not a quick study on mathematics and statistics (which I’m not – this stuff takes serious effort), you might be curious why adding up the average number of searches across all these verticals doesn’t yield the same number as the average for all Google searches. It took me a while to realize that’s because these averages are only for people who use those individual sections. E.G. If I didn’t perform any Google Video searches in a month, I wouldn’t be counted in the dataset of people who searched Google Video, but I’m still searching Google.com (default).
Next Steps for this Research
What we’ve got here is a remarkable baseline for how and how much Americans use Google. My hope for the future is to perform this study again in the Fall of 2025 to see how things have changed. Many suspect that AI tools will displace Google’s market share, and if so, we should see the number of searches per searcher per month fall.
We’ll also be able to spot when and whether other parts of Google search are changing in relation to one another, and whether AI or other competitors come for features like Maps, Shopping, News, or Images. My personal suspicion is that AI tools as a replacement for search is unlikely among most Internet users. A small group of passionate early adopters has definitely switched, but I think it’s unlikely we’ll see that penetrate the mainstream in the next year. Thankfully, this data, tracked over time, can help prove my hypothesis (and many other guesses) right or wrong. Stay tuned, because we’ve got that research on deck next week!
The data displayed in this report has been provided by Datos, A Semrush Company. The analysis is based on Datos’s US panel, representing a diverse and statistically significant sample of users, and covers the time frame of (01/23-01/25). For further information please visit Datos’s website and its Privacy Policy.