Understanding it
Google describes a technique called “query fan-out”: the question is broken into subquestions, run in parallel, whose results are assembled into a single answer. A page can therefore be picked for one specific aspect of the question, not only for its main query.
Unlike an AI Overview, which sits on top of the traditional results, AI Mode is a separate experience. Users stay there, ask one question after another and only see links as sources.
For a website, the logic stays the same as SEO: indexed pages, precise answers and proven expertise. Google requires no special optimization to appear there.
A concrete example
Illustrative example: a B2B buying journey
A marketing director asks: “which agency should we choose to redesign an insurer’s website, with accessibility requirements?” AI Mode searches in parallel for specialized agencies, accessibility obligations and feedback from past projects, then writes a summary.
An agency that publishes a precise industry page, detailed case studies and a page on accessibility covers several of these subquestions. It has a better chance of being cited than a site that only says “we build websites”.
The common mistake
Targeting only the main query
AI Mode assembles answers to subquestions. A page that repeats the main keyword without covering related questions (price, timelines, constraints, alternatives) leaves those pieces of the answer to other sites. Cover the questions surrounding your offer.