Zero-click search is not coming. It is here. In 2026, 68% of Google searches in the US end without a single click, up from roughly 60% two years ago. For comparison and review sites, the math has changed fundamentally: the page-one ranking you worked years to earn now drives a fraction of the traffic it once did.
What the Numbers Actually Say
SparkToro's 2026 panel study puts the zero-click rate at 68.01% for US searches. Other methodologies come in lower, around 58% to 65%, depending on the query mix and the panel. The spread matters less than the direction: every study points the same way.
Ahrefs tracked CTR across informational keywords before and after AI Overviews appeared. At position one, CTR fell from an average of 0.073 in December 2023 to 0.016 in December 2025. That is not a rounding error. One UK publisher tracked a page that kept its first-position ranking. CTR fell from 5.1% to 0.6%. Same position, same page, same content. The traffic redistributed above it.
Watch: Search YouTube for "zero-click search affiliate SEO 2025" — SparkToro and Ahrefs have published video walkthroughs on what this shift means for publishers and comparison sites.
What Died
Informational queries drove affiliate revenue for a decade. Someone searches "best casino bonuses UK," reads a comparison, clicks through, converts. That chain still exists, but the AI Overview now answers the informational half of the query without a click.
Template comparison pages are the most exposed. A page that lists ten casinos with a bonus column, a rating column, and a visit button is exactly what an AI Overview summarises. The user reads the summary. The click goes somewhere else, or nowhere at all.
Low-intent traffic built around programmatic SEO pages, "best X in Y city" formats, and thin listicles is structurally at risk. That business model produced page views for ad revenue. The page views are not coming back the same way.
What Survived
Transactional queries held. When someone searches for a specific brand or product, AI Overviews appear less than 5% of the time, and branded CTR can actually increase. People looking to sign up somewhere specific still click through.
Original research survives and gets cited. A site that publishes real data, primary sources, or analysis that cannot be found elsewhere becomes a source the AI Overview references. The citation still drives a smaller number of higher-quality visits.
Long-tail queries with specific intent also hold better. Someone searching "casino deposit limit change time UK" wants procedural information, not a summary. Complex, procedural, jurisdiction-specific content is harder for an AI Overview to compress into three sentences.
What to Build Instead
The sites gaining ground in 2026 are doing four things differently.
They publish original data. A site that runs its own tests, surveys its own audience, or tracks its own metrics has content that cannot be replicated by scraping. That content earns citations. Citations drive both direct traffic and brand impressions even when the click doesn't follow. Getting cited in AI Overviews is a skill you can build deliberately, even as a new site.
They build brand visibility off Google. Newsletter subscribers, YouTube audiences, and social followers are not affected by zero-click rates. A comparison site with 40,000 newsletter subscribers is not dependent on whether Google shows an AI Overview for "best casino bonus."
They own the transactional layer. The review article may get fewer clicks. The affiliate link on that article matters less than the conversion path on pages people arrive at with clear intent. Sites investing in landing page quality and conversion rate outperform sites still optimising purely for organic traffic volume.
They get cited, not just ranked. Appearing inside an AI Overview produces measurably higher CTR than ranking without a citation. The tactics that earn citations — answering questions directly with cited evidence, credentialed authors, named sources — are different from the tactics that historically earned rankings.
The Bigger Shift
Zero-click search is not an algorithm change you recover from. It is a structural change in how information moves. Google built a product that answers questions. It is good at it. Sites that exist to answer questions are now competing with Google itself.
The sites that survive are the ones that do something Google cannot: take a position, publish original findings, build a relationship with a specific audience that keeps coming back by choice rather than by search. Sites in YMYL categories — gambling, finance, health — face an additional layer of scrutiny that makes this adaptation more urgent, not less.
The traffic model that ran comparison and review sites for fifteen years is not dead. But it is smaller, and it is not coming back to what it was.
Related reading: