Position

How it is computed

Within each response, every mentioned brand receives a position based on the order in which the brands first appear in the answer text: the first brand named is position 1, the second is position 2, and so on. A brand's position metric is the average of its positions across the responses matching the active filter, rounded to one decimal.

position = round(average of the brand's per-response positions, 1 decimal)

Lower is better, and 1.0 is the best possible value. Only the first appearance in a response counts, so a brand named early and then repeated later still holds its early position. Responses that do not mention the brand at all are excluded from the average, and when the brand is not mentioned anywhere in the filtered period, position has no value.

Reading it correctly

Because unmentioned responses are excluded, position says nothing about how often you appear; that is what visibility measures. A brand can hold a strong average position of 1.5 while appearing in only a tenth of the answers. You should therefore read position together with visibility: visibility tells you how often you are in the answer, position tells you where in the answer you are when you make it in.

Position is most meaningful on prompts whose answers are lists, such as "what are the best X" questions. On a prompt whose answer discusses a single product in prose, nearly every mentioned brand is position 1 and the metric carries little signal. Filtering by tag keeps the list-style prompts separate.

Where it appears

Position appears on the project overview, is charted over time, and is listed per brand in the brand table. On the chart, rank 1 sits at the top, so an improving position moves a brand upward, matching the intuition that up is better.