Which Google searches are most likely to trigger one of those AI-written summary boxes up top?

This explores which kinds of Google searches are most likely to bring up an AI-generated summary. The short answer is that the corpus doesn't have a clear breakdown of which queries trigger AI Overviews. None of the notes here sorts searches by type, topic or wording and reports how often each one gets a summary. The nearest material is Pew's browsing-panel study of nearly 69,000 real Google searches Do AI summaries on Google reduce clicks to actual websites?, and Ahrefs' comparison of 'AI Overview keywords' with ordinary keywords How much are AI Overviews actually reducing organic search clicks?. Both studies start by splitting searches into ones that got a summary and ones that didn't. As summarized here, though, they report what happened after that split, not what caused it. If you want the trigger patterns, the Pew study is the best place to start.

The corpus is much stronger on what a summary does once it shows up, and the answer is fairly consistent. Pew found people clicked a result link 8% of the time when a summary appeared, compared with 15% without one, and the whole session was more likely to end with no clicks at all Do AI summaries on Google reduce clicks to actual websites?. Ahrefs estimates that clicks on the top-ranked result drop by about 58% for keywords that get a summary How much are AI Overviews actually reducing organic search clicks?. An eye-tracking study shows why: the summary now takes the attention the first result used to get. That result's share of attention fell from 31% to 9%, and people trusted the summary about as much as the ranked links Where do searchers look when AI Overviews appear?.

This lost traffic can be measured downstream. A natural experiment comparing English Wikipedia with its German and French editions found that AI Overviews cut search traffic to English Wikipedia by about 5% Does AI search summaries divert traffic away from Wikipedia?. News publishers expect their Google traffic to fall 40–43% over three years Will AI Overviews reduce search referral traffic to publishers?. A study from outside search explains the mechanism. When Nextdoor made its AI-written notification summaries more informative, fewer people clicked through, because the summary had already told them what they needed Does better summary writing actually increase user engagement?. The better the summary does its job, the less reason anyone has to visit the source.

The less obvious finding is about learning rather than traffic. Across seven experiments with over 10,000 people, those who learned a topic from an AI summary reported learning less and felt less ownership of what they knew. Their advice to others was also sparser than that of people who learned through ordinary web search Does learning from AI summaries produce shallower knowledge than web search?. So the question of which queries get a summary also decides which topics people now learn about in a shallower way. People haven't stopped using regular search, though. Nielsen Norman Group found users switching between AI chat and traditional search throughout their tasks Does generative AI chat actually replace traditional search?.