The Click Reality: How Much Traffic AI Overviews Eat
Google's AI Overviews have been live for just over a year, and the data is finally thick enough to see a pattern. Across 51,000+ tracked events, the picture that emerges is less of a gentle evolution and more of a structural shift. For the top-ranked result, clicks are down by 58%, up from 34.5% in April 2025. That means for every 100 clicks a page used to earn, Google now keeps 58. That is a staggering number, and it arrives with a side of volatility: any given AI Overview has a 70% chance of changing between observations, with cited content lasting about 2.15 days on average. Nearly half — 45.5% — of cited sources are entirely new each time. If you check once and feel confident, the data says you're probably looking at stale results.
The Ahrefs AI Overviews Tracker, which runs real queries through Google and stores responses in Brand Radar, also points to brand mentions as the single strongest link to AI visibility. Across 75,000 brands, the more the web mentions you, the more AI Overviews surface you. That means digital PR, roundups, and "best of" lists aren't just vanity metrics — they're visibility factors.
Volatility and Source Turnover: The 70%/45.5% Problem
Volatility is the defining characteristic of AI Overviews. The Ahrefs data confirms what many SEOs have suspected day to day: a 70% chance the overview will look different on the next run. Content persists for roughly 2.15 days, and almost half the citations are fresh each observation. That speed means a one-off check goes stale fast. For brands trying to monitor their citation status, periodic tracking is the minimum, not the exception.
The Search Engine Journal pullback study adds a parallel dimension. After the PR backlash that followed early harmful AI Overview answers, Google cut AIOs by two-thirds across almost every industry. The share of queries showing an AIO dropped from 42% to 23, almost a halving. That reduction may feel like good news for organic defenders, less real estate ceded to the AI summary, but the flip side is that citations per AIO doubled, from 15 to 32. More links in a smaller slice of queries changes the math.
Citations Increase, But Who Gets Them?: Reddit, Wikipedia, and the Top-Ranked Effect
One of the most consistent findings across both data sets is the dominance of the No. 1 organic result. Both the Ahrefs tracker and the Search Engine Journal analysis show that AI Overviews lean heavily on the top web result for text snippets. When TIME ranks No. 1 for "best hair growth products," the AIO text mirrors that article's structure, even when the product list inside the AIO changes. Google seems to use Retrieval-Augmented Generation (RAG) to pull fragments from the highest-ranked page, then reorganize them with citations.
After the backlash, Reddit and Wikipedia citations both increased. In the SEJ sample of 1,675 keywords, Reddit appeared in more than 10% of citations for only 28 keywords after the pullback, and Wikipedia for 28 as well. But the trend is clear: both UGC domains are being leaned on more heavily. The Ahrefs data agrees, listing "Show up on UGC sites" as one of six factors that help a brand show up in AI Overviews. Being active on Reddit and Quora matters because Google's systems trust those platforms.
Industry Impact: YMYL in the Crosshairs
The pullback was not uniform. YMYL, Your Money, Your Life, industries were the hardest hit. Health, science, and law saw the largest drops in AIO frequency. Meanwhile, a few sectors picked up a negligible 0.2% more AIO visibility, but that is noise at the margin. The message is that Google is cautious with sensitive topics, and brands in health, science, or legal spaces should expect both lower AIO frequency and stricter scrutiny when citations do appear.
What Brands Can Do: From Mentions to Visibility
The data from both sources converges on a few practical takeaways. First, brand mentions are the strongest visibility signal we have. The Ahrefs study of 75,000 brands makes that link explicit: more web mentions = more AI Overviews. Second, UGC presence matters. Being genuinely active on Reddit, Quora, or YouTube (Google partners with Reddit) gives your brand a better shot at citation. Third, keep content fresh. Gemini, which powers AI Overviews, prefers newer pages. Use tools like Site Explorer to identify declining pages and update them on a cadence. Fourth, build long-tail clusters. AIOs show up most for informational, longer-tail queries, so covering topic clusters broadly increases the chance you're an answer for more of them.
The days of assuming organic rank alone determines AI citation are over. Classic search performance is necessary but no longer sufficient. The brands that will thrive are the ones investing in mentions, UGC, freshness, and cluster coverage, not just keyword optimization.
Conclusion: The New Normal for Organic Search
AI Overviews are here to stay, and their structure will keep shifting. The 58% click reduction for top results, the 70% volatility rate, and the growing citation count all point to a search landscape where organic visibility is no longer a single-number game. The brands that treat AI Overviews as a distinct channel, with its own metrics, its own optimization levers, and its own content cadence, will be the ones that keep earning traffic. The rest will keep checking their rankings and wondering why the clicks keep shrinking.
*Data sources: Ahrefs AI Overviews Tracker (brand mention correlation, CTR reduction, volatility metrics); Search Engine Journal "AIO Pullback" study (query frequency drop, citation count increase, YMYL industry impact). Verified source URLs: https://ahrefs.com/ai-overviews-tracker, https://www.searchenginejournal.com/aio-pullback-google-shows-2-3-fewer-aios-and-more-citations/521110/
