AI visibility: the number your brand does not know it has

Ask an assistant the question your buyer asks: best tool for [your category]. Either your brand is in the answer or a competitor's is. AI visibility is that fact, measured. It is becoming the front door of buying research and most brands have never once checked theirs.

The two numbers that matter

1. Named when asked directly. Ask the model about your brand by name. Does it describe you accurately, guess or draw a blank? This measures whether the models know you exist.
2. Named when asked to recommend. Ask category questions with no brand mentioned. This is share of voice: of all the answers, how often are you in them? This measures whether models send you buyers.

The gap between the two is the whole story. A brand described accurately when named but absent from every recommendation is known and not suggested and closing that gap is the work.

How to measure it honestly

• Fix a panel of buying questions, phrased the way buyers phrase them.
• Ask them across several models, not one. Answers differ by model and by buyer persona.
• Repeat the identical panel on a schedule. Movement only means something against a fixed baseline.
• Record who IS named when you are not. Your visibility problem is specific competitors owning specific answers and the fix starts with knowing which.

What moves the number

Models compose answers from sources: roundup pages, comparison pages, reddit threads, directories. Getting into those sources is the discipline covered in LLM SEO. It moves in weeks, not days, because models refresh their view of the web on their own schedules. The compounding is real: each source you enter makes the next citation more likely.