How to measure your brand's visibility in AI answers

A brand’s visibility inside AI answers is measured by how often and how prominently it gets named when engines — ChatGPT, Gemini, Perplexity, and others — answer questions in its category. The unit isn’t search rank; it’s presence inside the answer text. The basic method is simple: gather the questions buyers actually ask, put them to several engines, and record when the brand appears, in what position, and alongside whom.

Step 1 — build the right question set

Start from questions that lead to a buying decision, not from the brand’s own name. For example: “best provider for category X,” “recommended Y service in Jakarta.” These are the questions that decide whether a brand makes the shortlist at all. Important rule: never insert the brand’s name into a tracked question — that biases the result.

Step 2 — ask several engines

Answers differ across engines and change over time. Test each question on several major engines, and repeat on a schedule. A single test is only a snapshot; useful measurement is repeated so trends become visible.

Step 3 — record three things

For each answer, note: whether the brand is named, its position among the names that appear, and which other names show up with it. Those three notes turn impressions into data. From them, patterns emerge: where the brand is strong, where it’s absent, and which competitor most often takes its place.

Step 4 — roll it into one measure

The spread of findings can be summarized into a single visibility index — a blend of how often the brand appears, how high its position is, and how much weight the questions where it appears carry. That single number makes tracking over time easy: rising means the brand’s narrative is earning more machine trust; falling means the opposite.

What to avoid

Don’t measure by naming the brand directly to an engine (“what is brand Z”) — that tests memory, not recommendation. And don’t conclude from a single answer; models are probabilistic and results vary. Valid measurement uses neutral questions, multiple engines, and repetition. With this foundation, AI visibility stops being a guess. It becomes something you can track, compare month over month, and treat as a new performance indicator alongside traffic and impressions. For the single metric this rolls up into — and why it replaces search rankings as a KPI — see Visibility Index: the KPI that replaces search rankings.

White Wood is a Narrative Architecture studio in Jakarta that helps brands get named and recommended inside AI answers.