By Stuart Kerr, Technology Correspondent, LiveAIWire
Publishers survive the zero-click era or they don’t, and the data on which way that’s trending has gotten considerably sharper since Google’s AI Overviews and AI Mode began reshaping search. A joint Datos and SparkToro clickstream study, tracking millions of real searchers rather than sampled keywords, found that for every 1,000 Google searches in the US, only 360 result in a click to the open web. In the EU, that figure is slightly higher at 374. Put another way, roughly two-thirds of all Google searches in 2024 ended without sending a single visitor anywhere outside Google’s own ecosystem.
The picture sharpened further in July 2025, when analytics firm Authoritas submitted research to the UK’s Competition and Markets Authority as part of a formal complaint against Google, finding that a site previously ranked first in search results could lose approximately 79 percent of its traffic for that query once an AI Overview appeared above the link. Carly Steven, an executive at MailOnline, told The Guardian the site was seeing clickthrough rates fall by 56.1 percent on desktop and 48.2 percent on mobile specifically on searches where an AI summary appeared.
Why Google Disputes the Numbers
Google has publicly rejected the framing of these studies. A company spokesperson called the Authoritas and Pew research inaccurate and based on flawed assumptions, arguing the underlying data was outdated and unrepresentative, and stating that Google continues to send billions of clicks to websites every day without evidence of dramatic aggregate traffic decline. That dispute is genuinely unresolved. What both sides broadly agree on is that individual search result pages featuring an AI summary see measurably lower click-through than pages without one. The disagreement is over what that means once averaged across Google’s entire search volume.
What This Means for How You Structure Content
For any publisher, the practical response starts with accepting that being cited inside an AI summary and being clicked are now two separate outcomes, not the same one. Content structured with clear, extractable answers, proper schema markup, and specific, well-attributed facts is more likely to be the source an AI summary actually cites, even if that citation converts to a click far less often than a traditional blue link once did. Owen Meredith, chief executive of the UK’s News Media Association, has called the current arrangement unsustainable, arguing that Google is monetising content it did not create while making it structurally harder for the outlets that created it to reach the readers who fund their work.
The Legal and Regulatory Fight Underway
The Authoritas research was submitted as part of a joint legal complaint to the UK’s Competition and Markets Authority, filed by the tech accountability group Foxglove, the Independent Publishers Alliance, and the Movement for an Open Web. Foxglove director Rosa Curling has argued the practice goes beyond simple content use, framing it as Google using publishers’ work to fuel its own tools and profits while making it structurally harder for those publishers to reach their audience. The CMA has not yet issued a public ruling. Separately, in the US, Bain and Company’s own consumer research found roughly 80 percent of people now rely on AI-generated summaries for at least some portion of their searches, a shift the firm frames as a fundamental redefinition of what search marketing needs to accomplish, rather than a temporary dip to wait out.
The Diversification Strategy Publishers Are Actually Trying
Beyond content structure, the more durable response several publishers are pursuing is reducing dependence on search referral traffic entirely. Owned channels, email newsletters, paid subscriptions, apps, and direct community platforms, do not depend on how a search engine chooses to summarise or cite a given page. Some larger publishers have also begun negotiating direct licensing arrangements with AI companies for training data access, though smaller outlets generally lack the leverage to negotiate similar terms. None of these approaches fully replaces the traffic volume search once reliably delivered, but each removes at least one point of dependency on decisions publishers do not control.
The honest answer to whether publishers survive the zero-click era is that some already are not, and the ones with the strongest owned-audience relationships are faring measurably better than those relying primarily on search referral. As our own reporting on AI’s invisible infrastructure and the silent bias built into automated curation has explored, the tools deciding which content gets surfaced and cited carry real editorial weight, even when they are marketed as neutral search results rather than a curated product. Whether regulators in the UK, EU, or US eventually force a more balanced arrangement, or publishers simply adapt to a smaller, more competitive slice of referral traffic, is the question the next eighteen months of CMA and antitrust activity should start to answer. For more on how this dynamic plays out at the search-interface level itself, see our coverage of AI Mode’s UK launch and the click-through data behind it.
About the Author
Stuart Kerr is Technology Correspondent at LiveAIWire, covering artificial intelligence, emerging technology, and their impact on business, society, and everyday life. LiveAIWire publishes original AI journalism every weekday at liveaiwire.com.