When Google launched AI Overviews in May 2024, the company framed the feature as a helpful addition to search, designed to give users faster answers to their questions. What followed was a dramatic collapse in traffic to content creators and publishers worldwide. The data tells a devastating story: some of the largest media outlets in the world have lost between 30 to 97 percent of their organic traffic from Google searches in less than two years.
The numbers are staggering. A detailed analysis by SEO firm Growtika tracked traffic to 10 major tech outlets from early 2024 to early 2026. These publications collectively received 112 million site visits per month from Google users at their peak. By January 2026, that number had plummeted to under 50 million—a decline that accelerated dramatically in mid-2025 when Google expanded AI Overviews to cover a much broader range of queries. Some outlets experienced apocalyptic drops: Digital Trends went from 8.5 million clicks monthly to just 264,861—a 97 percent collapse in web traffic.
The Zero-Click Search Crisis
The core problem is what researchers call “zero-click searches.” When Google’s AI Overviews appear at the top of search results, users no longer need to click through to publisher websites. They get their answer directly on Google’s page. Data from Similarweb showed that news searches resulting in zero clicks—where users didn’t click any links—rose from 56 percent in May 2024 to nearly 69 percent by May 2025. A Pew Research Center study found that only 1 percent of users clicked on links included within AI Overviews, compared to 15 percent of users who encountered traditional search results.
This shift represents a fundamental change in how search works. For two decades, Google’s business model depended on sending traffic to publishers. Publishers created content, Google indexed it, displayed it in search results, and earned advertising revenue from the search page itself. Publishers, in turn, earned money from the traffic Google sent them. This symbiotic relationship is now breaking down. When Google synthesizes answers directly, no one has reason to click through anymore.
Measuring the Damage Across Industries
The impact varies significantly by content type and industry. Informational and lifestyle content has been hit hardest—exactly the kind of material that AI Overviews handle most effectively. Mail Online reported that when AI Overviews appeared for its content, click-through rates dropped by more than 50 percent on desktop and 48 percent on mobile. CBS News saw 75 percent of its top 100 keywords that triggered AI Overviews result in zero clicks, compared to 54 percent for the same keywords without an AI Overview.
Food bloggers have experienced catastrophic declines. One creator saw 80 percent of her traffic and revenue disappear in two years. A travel blog that had operated successfully for over a decade lost half its traffic within months after Google launched AI Overviews and was forced to shut down entirely. Recipe developers describe AI-generated versions of their work as “Frankenstein recipes”—mashed-together combinations of ingredients from multiple sources, often inaccurate and missing crucial details like proper cooking temperatures and times.
Google’s AI Overviews Appear on Half of All Searches
The prevalence of AI Overviews has grown exponentially. In January 2025, these summaries appeared on just 6.5 percent of queries. By July 2025, AI Overviews peaked at appearing on nearly 25 percent of all Google searches. While coverage pulled back slightly to around 16 percent by November 2025, the feature now appears on roughly half of searches when counting multiple variations and regional differences. In healthcare, education, and B2B technology sectors, AI Overviews now trigger on 80 percent or more of queries.
This expansion matters because it determines which sources Google selects to create its summaries. The company pulls from multiple websites and synthesizes their content into a single answer. Publishers whose content gets selected receive a tiny amount of additional visibility—they’re cited as a source—but lose far more traffic from users who no longer need to click. Being cited in an AI Overview has become like being quoted in a news article: it’s brand awareness without business value.
The Attribution Problem and Disappearing Revenue Streams
Beyond zero-click searches, publishers face another challenge: AI Overviews often show links to source material, but users almost never click them. Google’s own CEO claimed in interviews that content appearing within AI Overviews gets higher click-through rates than when placed outside, but independent research contradicts this. The reality is that users trust the AI summary and have no reason to verify by clicking original sources.
This affects multiple revenue streams simultaneously. Publishers relying on affiliate marketing—earning commission when readers click links to purchase products—have seen affiliate revenue drop 20 to 40 percent. Advertising revenue has declined because fewer visitors reach publisher sites. Subscription models suffer when casual readers get summaries instead of being drawn into paywalled content. Some publishers have had to lay off significant portions of their staff or close entirely as traffic collapsed.
Who Benefits? Wikipedia and Reddit Dominate AI Summaries
Interestingly, certain sources consistently appear in AI Overviews far more than others. Wikipedia and Reddit dominate citations in both Google AI Overviews and ChatGPT outputs. Wikipedia appears in roughly 45 percent of top domain citations across datasets, while Reddit accounts for nearly 30 percent. This creates a strange dynamic: Wikipedia benefits from being cited without traffic consequences since people don’t click to verify encyclopedia entries. Reddit sees increased traffic as users follow links to read comments and discussions. Meanwhile, news publishers and independent creators see their content summarized and their traffic evaporated.
Google’s algorithm favors certain source types. Google AI Overviews lean heavily toward news-driven sources and controversy indexing, while ChatGPT pulls more from product reviews, forums, and social discussions. A study by BrightEdge found that Google’s AI Overview is 44 percent more likely than ChatGPT to surface negative brand sentiment overall. ChatGPT concentrates its criticism 13 times more heavily near the point of purchase. Even worse, the two systems disagree on which brands to criticize 73 percent of the time.
Regulatory Pushback and Licensing Negotiations
Publishers have fought back through lawsuits, regulatory complaints, and licensing negotiations. The UK’s Competition and Markets Authority has launched investigations into whether Google is abusing its monopoly position in search by forcing publishers to provide content for AI Overviews without adequate compensation. Google has responded by proposing to pay for “access” to content through licensing deals, but refuses to pay for training data—a distinction that publishers consider meaningless.
Google clarified its position in testimony to UK lawmakers: it will pay for access to content but not for training. The company argues that training large language models on unpaywalled content represents “statistical analysis” rather than copying, and therefore should not require payment. Publishers view this distinction as false—if Google can generate revenue from content trained on publisher material, the creators of that material deserve compensation.
The Path Forward: Publishers Adapt or Disappear
Facing this new reality, publishers are adopting survival strategies. The most successful approach involves building direct relationships with audiences through newsletters, subscription services, membership programs, and owned platforms. Publishers are moving away from reliance on Google traffic and investing in channels they control directly. Some are investing heavily in video content, which still drives meaningful traffic. Others are doubling down on investigations, original reporting, and exclusive content that AI cannot easily replicate.
Publishers are also optimizing content structure for AI visibility. Putting key answers at the top of articles, using clear subheadings that pose questions, implementing structured data markup, and creating FAQ sections all increase the likelihood of being cited in AI Overviews. Being featured in an overview, while not driving clicks, builds brand authority and can create halo effects that improve performance across other channels.
The media industry faces an existential threat as Google transitions from a search company to an answer company. Publishers who built their business models entirely around Google traffic are discovering too late that they were renting an audience, not owning one. Those who invested in direct relationships, distinctive editorial voices, and owned distribution channels years ago are surviving. The rest are adapting quickly or disappearing.
WebProNews is an iEntry Publication