Google's experiment shows news content has minimal impact on search revenue
Analysis of EU test reveals publishers may need Google more than Google needs publishers.
Google's recently concluded European experiment has revealed that removing news content from search results had minimal effect on the company's user engagement and revenue. The test, which ran from November 2024 to January 2025, has significant implications for the ongoing negotiations between tech platforms and news publishers regarding content compensation.
On March 21, 2025, Google published results from its European news content experiment that began four months ago. According to the data released by Google, removing news content from search results led to only a slight decrease in user visits, with search ad revenue remaining essentially unchanged.
The test, which started on November 14, 2024, involved 1% of users in eight European countries: Denmark, Belgium, Spain, Italy, Greece, the Netherlands, Poland, and Croatia. For these users, Google removed results from press publications that were eligible under Article 15 of the European Copyright Directive (EUCD) across Search, Discover, and Google News platforms.
According to Paul Liu, Director of Economics at Google and author of the report, "The data showed that Search ad revenue did not change despite daily average users (DAUs) declining by 0.8 percent, which is consistent with users continuing to use Google for more commercial queries even as they used it less for news queries."
The experiment occurred against the backdrop of increasing pressure on tech platforms to compensate news publishers for content used in search results. The EU Copyright Directive, passed in 2019, requires that Google and other tech platforms compensate publishers for using their content for anything other than "very short snippets."
Key findings
The detailed report released by Google contained several notable findings:
- Search Daily Active Users (DAUs) declined by 0.77% when news content was removed
- Discover DAUs saw a more significant drop of 5.47%
- Google News DAUs showed no statistically significant change
- Search ad revenue showed no statistically significant change despite the user decline
- Discover revenue declined by approximately 2%
- The overall ad revenue impact across all Google properties could not be statistically distinguished from zero
The experiment methodology involved removing 13,409 domains identified as "press publications" within the meaning of Article 15 of the EUCD. Google's experimental infrastructure confirmed that impressions and clicks for these affected domains declined by 96-99% across all three tested properties.
Sulina Connal, Managing Director of News and Publishing Partnerships at Google, had announced in November 2024 that the experiment was being conducted to "assess how results from EU news publishers impact the search experience for our users and traffic to publishers." The February 4, 2025 update confirmed the experiment's conclusion, noting that Google had "collected sufficient data."
The broader context
The experiment comes at a critical time in the relationship between Google and news publishers. Since 2019, the EU Copyright Directive has established a framework requiring tech platforms to compensate publishers for using their content. Google has implemented its Extended News Previews (ENP) program, signing agreements with more than 4,400 publications across 24 European countries.
Similar legislative efforts have emerged in other jurisdictions. Google and other tech platforms face pressure in Canada, Australia, and elsewhere where publishers are seeking compensation for content use. The experiment results may influence these ongoing negotiations.
Alan Chapell, writing for The Monopoly Report on April 2, 2025, characterized Google's experiment as "simply part of a larger negotiating strategy" in these discussions. "On some level, Google's experiment is simply part of a larger negotiating strategy. The EU Copyright Directive (2019) requires that Google and other tech platforms compensate publishers for using their content for anything other than 'very short snippets'," Chapell noted.
The experiment follows a period of significant changes to Google's search results, which have increasingly kept users within Google properties. Elizabeth Reid, Head of Google Search, told Bloomberg in a profile published on March 25, 2025, that the Google search bar will become less prominent over time as voice and visual searches grow in importance. This shift represents a fundamental change to Google's traditional role of directing users to external websites.
A former Google senior executive quoted by Bloomberg characterized the company's historical relationship with publishers bluntly: "Giving traffic to publisher sites is kind of a necessary evil. The main thing they're trying to do is get people to consume Google services."
Historical parallels
Industry analysts have drawn parallels between Google's current approach to publishers and historical moments of economic tension. Chapell compared the situation to New York City's fiscal crisis in the 1970s, when President Gerald Ford initially refused federal assistance to the struggling city, prompting the famous "Drop Dead" headline in the New York Daily News.
"The reason I think the Ford analogy is apt is that President Ford seemingly forgot how valuable an economic engine NYC had been and could be. Similarly, Google seems to have forgotten how its economic engine has long been driven by publishers," Chapell wrote.
This perspective stands in contrast to statements made by Google co-founder Sergey Brin in 2009, who emphasized the symbiotic relationship between Google and publishers: "It's very important to Google that all these websites [and] information sources do well, thrive and continue to create information. Without it there would be nothing to search."
Implications for publishers
The experiment results suggest that news publishers may be more dependent on Google than vice versa. With Search Daily Active Users declining by less than 1% and no significant impact on search ad revenue, Google appears to have demonstrated that news content is not critical to its core business model.
For publishers, these findings are concerning. The experiment indicates that Google could potentially abandon news content entirely without significant business impact. This strengthens Google's negotiating position in discussions about compensation for news content.
The country-level analysis provided by Google showed consistent patterns across all eight test markets. Web Search DAUs declined slightly in all countries, while Discover DAUs saw more substantial drops ranging from about 4% in Poland to nearly 10% in Belgium. Revenue impacts varied by country but remained statistically insignificant overall.
Technical impact assessment
Google's experimental methodology was robust, employing a randomized controlled trial (RCT) that maintained consistent user groupings throughout the test period. This allowed for observation of both immediate and longer-term behavioral changes as users adapted to the absence of news content.
To validate the experiment, Google first ran a "no-op" control test from July to November 2024, confirming that the experimental infrastructure was properly configured. During the actual experiment, the removal of eligible content was nearly complete, with impressions and clicks for affected domains declining by more than 96% across all three tested properties.
The technical report included detailed metrics on Search DAUs, Discover DAUs, Google News DAUs, and corresponding revenue impacts. Statistical analysis confirmed that most metrics had stabilized by January 2025, allowing Google to end the experiment with high confidence in the results.
Critical perspectives
Some industry observers have questioned Google's motivations and the broader implications of the experiment. Chapell raised several concerns about Google's strategy:
- Providing evidence that Google can "degrade" its service without revenue impact undermines arguments against monopoly claims
- The tracking capabilities demonstrated in the experiment raise privacy questions
- The short-term focus fails to account for the longer-term societal value of journalism
- The experiment demonstrates Google's ability to effectively "turn off" access to news, raising concerns about information control
"So yeah, this experiment may serve as a data point in the negotiation with publishers and other content creators. But I think Google will be hard pressed to find many publishers, regulators or policymakers who view it as a credible data point," Chapell wrote.
Why this matters to marketers
For marketing professionals, Google's experiment has several significant implications. The findings suggest that news content plays a relatively small role in Google's overall search ecosystem, particularly from a revenue perspective. This may lead to shifts in how content is prioritized within search results.
Marketers should note that users appear to maintain their commercial search behaviors even when news content is removed. This reinforces the distinct intent behind different types of searches and suggests that marketing strategies focused on commercial search terms may be insulated from changes to news content presentation.
The experiment also highlights the growing tension between platforms and content creators in the digital ecosystem. As these relationships evolve, marketers will need to adapt their strategies to ensure visibility in an environment where the rules of engagement between platforms and publishers are being renegotiated.
The results also come as Google continues to implement other significant changes to its search interface, including AI Overviews and a reduced emphasis on the traditional search bar, according to statements from Google's Head of Search Elizabeth Reid. These changes collectively represent a fundamental shift in the search landscape that marketers must navigate.
Timeline
- 2019: European Union passes the Copyright Directive, including Article 15 requiring platforms to compensate publishers
- November 14, 2024: Google begins experiment in eight European countries removing news content for 1% of users
- January 31, 2025: Experiment concludes after metrics stabilize
- February 4, 2025: Google announces end of experiment and plans to analyze data
- March 21, 2025: Google publishes experimental results showing minimal impact on search usage and revenue
- March 25, 2025: Bloomberg publishes profile of Google Search head Elizabeth Reid discussing reduced prominence of traditional search
- April 2, 2025: The Monopoly Report publishes analysis criticizing Google's approach to news publishers