Loading…

Documentation

The product reference: how the measurement is put together, where to start and what each module does. The normative part (exact parameters, formulas and limits) lives in the methodology.

How the measurement is put together

  1. 1. Entry: the scanner

    A one-off measurement of 10 responses, free and without sign-up, returning a 0 to 100 score with a published formula. It tells you where you stand and whether measuring seriously is worth it.

  2. 2. Setup: brand and prompts

    You register the brand with its aliases, domains and competitors, and approve the prompts to be measured, each with its language and country. Nothing enters the measurement without your sign-off.

  3. 3. Measurement: sweeps through official APIs

    Each sweep runs the approved prompts on your plan’s engines, with web search in the state we declare and repetitions to stabilise the sample. The full response is stored, not a summary.

  4. 4. Reading: metrics, evidence and alerts

    Responses are aggregated into metrics with visible denominators, every figure drills down to its evidence, and alerts only fire with statistical testing.

Where to start depending on your goal

  • Know where I stand today

    Run the scanner: about 60 seconds, no sign-up, and a 0 to 100 figure with its breakdown in front of you.

    Free scanner

  • Measure weekly and compare

    Register your brand and approve your prompts. The trial lasts 14 days, asks for no card and has the Pro plan features.

    Pricing

  • Defend a figure in front of a client

    Read the methodology and drill from the figure down to the evidence: every percentage carries its denominator and its responses.

    Methodology

Reference by module

Each module has its page with what it does, how it works, what you see on screen, who it is not for and its availability by plan.

Documentation: how AI visibility is measured | Citenza