Google Test Shows Content Has No Measurable Impact On Ad Revenue
- by Laurie Sullivan @lauriesullivan, March 21, 2025
Google has been testing the value of European news content on its advertising revenue.
The test was done in response to a “number of inaccurate reports that vastly overestimate the value of news content to Google,” wrote Paul Liu, Google director of economics, in a blog post.
The test found no measurable impact from European news on advertising revenue in the eight European countries the company tested.
Google had to remove news content from more than 13,000 EU publisher domains for 1% of users in Belgium, Croatia, Denmark, Greece, Italy, the Netherlands, Poland and Spain. The test ran from November through January.
It removed content from Discover and Google News, and included users who had logged in with a Google user ID, and those who logged out with browser cookies.
The study showed that when Google removed content, there was “no change to Search ad revenue and a 0.8% drop in usage, which indicates that any lost usage was from queries that generated minimal or no revenue,” Liu wrote.
advertisement
advertisement
Google calculated the revenue impact for the overall "Google ecosystem,” consisting of all properties including Shopping and Display Ads. Different parts of Google's business will have different margins, so the contribution of each part to Google's bottom line differed.
For this calculation, Google applied a 77.3% gross margin for Search and Search-adjacent properties — Discover, AdSense for Search, tools related to geospatial data like Maps, and Shopping — and a 32% gross margin for Display Ads to come up with an overall revenue impact.
Google’s ecosystem impact did not indicate a statistically significant decline for either the logged-in subset or the population as a whole.
The study found that combined ad revenue across Google properties, including its ad network, also remained flat. The effects were tested on first-party sites like YouTube and Gmail as well as third-party sites that use its technology.
Google Discover saw a revenue decline of a statistically significant 2%, but the company said this aggregator service makes only a modest contribution to its overall revenue.
This study showed people come to Google for many other types of tasks, even when Google is less useful for catching up on news.
A report that outlined the entire test also notes agreement made with publishers in 2019, when the European Copyright Directive (EUCD) was passed, including Article 15, the “neighboring right” for press publishers.
The Directive outlines two important guiding principles. It notes that people and platforms can continue to link to, and include, very short extracts of publishers’ content.
At the same time, it created new rights for news publishers when extended previews of their work are used online.
So hmm. I don't quite understand the details, but, if I understood this correctly, what Google's internal report mostly says is that, when 13,000 news media web pages have been prevented from being seen by 1% of the inhabitants of 8 European countries... overall ad revenue, on average, only dropped by 0.8%; and, from that, they conclude that people don't go to Google for reading the news.
There are a few logical gaps somewhere in that reasoning.
Now, I understand that Google will never reveal the details of their internal reporting; after all, it contains considerable corporate secrets. It's not as if such reports can be published in peer-reviewed academic journals, either, so it's hard to prove or disprove them: we're simply not going to get the 'raw data' in order to confirm Google's calculations and conclusions. We just need to take their word for it. And, I suppose, it's better than the alternative, i.e., Google not even making an attempt to share their data with us.
Nevertheless, as said, even without knowing every detail, it's worth asking a few pertinent questions, namely, how was that conclusion reached, and what exactly would be the conditions for not reaching that conclusion.
Consider the following: there are billions upon billions of sites out there. Google 'only' removed 13,000. It seems quite a lot, but... how much revenue do they earn from those 13,000? And compared with... what? YouTube, which is, these days, Google's 'cash cow'? A single site which might be responsible for, say, half the ad revenue that Google gets? (I'm speculating, it might be even more than that — or not.)
Let's say that Google would remove 13,000 websites related to something else, such as, say, click-bait websites. Would these produce a similar income reduction — or would the reduction be more substantial? And why should it matter? Wouldn't the results be pretty much the same if they randomly removed 13,000 other websites, not related to news?
Sure, I can imagine that some kinds of sites are far more relevant than others. It's not only YouTube, of course; it's also travel & cooking websites. Or sports-related forums. Or... well, I don't know. You tell me where people spend time these days outside the so-called social media. How relevant, in the grand scheme of things, is the revenue got from that?
We'd have to see a breakout of revenue per class of content, or per industry, or whatever metrics Google uses to classify their sources of revenue. We would also need to understand how many people — even if just a percentage, not absolute numbers — are employed by Google to sell ads on the news media (or, rather, on sponsored ads when someone searches for something on the news media). Assuming they even organise their sales force that way.
We also don't know what 'a drop of 0.8%' means. Suppose that the total revenue from those 13,000 websites — which Google now claims that 'nobody sees' — are peanuts: just a few hundreds of Euros per month. Now they earn 0.8% of that. That's pretty much irrelevant — it won't even pay a cup of coffee for one member of the team doing the research. 0.8% less revenue from a block that already provides with next-to-nothing revenue is not only statistically insignificant, but maybe Google shouldn't even bother indexing those guys. We don't know!
Google's role as the dominant search engine is sustained by having crawled the most content (since they started first, long, long ago...) and having the arguably best search facilities on top of that data. They managed to give the world the notion that 'Google crawls everything, Google knows everything'.
Therefore, these worries are more related to the image projected by Google, and less by the subjacent reality of how Google's revenues are calculated — something they will not share with us, of course.
Crazy, right? After all this time Google has been saying the opposite.