What is Share of Answer?

In ranked search, ten results share a page and position is a spectrum. In generated answers, presence is closer to binary. A brand is either in the synthesis or it does not exist for that query. Ahrefs' study of roughly 300,000 keywords found top result click-through collapsing from 7.3 percent to 1.6 percent where AI Overviews appear. The click moved into the answer, so measurement has to move there too.

A rigorous program has four components. First, a defined prompt set: the high intent questions of a category, weighted by commercial value. Second, repeated sampling across models, because answers are probabilistic and a single query is an anecdote. Third, scoring for presence, position, and sentiment: named at all, named first, recommended or merely mentioned, described accurately. Fourth, longitudinal tracking, because the movement is the signal. A static score is a snapshot. A time series is intelligence.

Two adjacent metrics complete the picture. Citation share measures how often a brand's own domain is the cited source behind an answer. Recommendation share measures how often the brand is the model's actual pick rather than a passing mention.

The current state of practice leaves room to lead. Per McKinsey, only about 16 percent of brands systematically track AI search performance, and methodology across the young tooling category remains unstandardized, which is precisely why published, auditable measurement matters.