AI Overviews

AI Overviews matter for two reasons: reach and click absorption. They reach about 2 billion monthly users (DemandSage, from Google I/O disclosures), appear on somewhere between 20% and 43% of searches depending on the study, and when present, click-through to organic results drops roughly 60% (Search Engine Land / SparkToro). For many businesses this is the AI surface their customers hit first, often without knowing they're using AI at all.

Mechanically, AI Overviews are built from Googlebot's normal index — there is no special file or crawler to optimize for. That makes the levers familiar: standard crawlability, passage-level clarity, and structured data. A page that answers a question directly in a liftable paragraph can be cited in an overview even when it doesn't rank first in the classic results below.

Example: a first-time diabetes patient searches for the best diabetes management app in India. The overview synthesizes a shortlist from three app-review sites and clinical directories, and cites them. Those three sites get the click-throughs and the credibility; every other brand page is beneath an answer most users never scroll past. Related but distinct: Google AI Mode, its logged-out conversational search surface, which is growing fast from a small base. We track AI Overviews and AI Mode alongside ChatGPT and Perplexity because their selection logic differs measurably — see citation share.