Skip to content

Methodology

How our AI Visibility Scan works

Our Signal Report distinguishes website evidence from untested AI answers and booking tasks. Every number we report is a count of something we observed, and we say what was counted.

  1. 01

    Our current website scan

    We review public website access, business information, contact signals, and booking links. We do not test live AI answers, place calls, submit forms, or complete bookings. We document findings and unchecked items in a Signal Report.

  2. 02

    Our live answer-check has not run yet

    We have not run our live AI-answer check. Once available and separately agreed, we will run customer questions and save the actual answers, platform, and date. We will distinguish business names from links to sources and will not present website findings as AI-answer evidence.

  3. 03

    Our booking checks have limits

    We review visible contact links, forms, and booking paths. Finding a link does not establish successful delivery or a completed booking. Live AI-agent testing requires separate authorization and availability; it is not included in our automated website scan.

  4. 04

    Our work and re-check

    We agree the eligible corrections, price, and delivery schedule in writing before starting the Fix. We complete the authorized work and document our re-check in a Signal Report. We cannot promise what an AI assistant will say, recommend, or book.

  5. 05

    How tests are run.

    • We report counts and actual evidence, without scores or percentages.
    • We identify the sources and dates of completed observations.
    • We mark missing evidence as unavailable or not checked.
    • We require prior written authorization for live calls, forms, or booking tests.

    The automated scan makes no calls, sends no messages and completes no bookings. Any separate task testing requires prior authorization.

Accountability

What we promise, and what we don't.

What we count, and where.

Automated website scan
Observed website checks only. No visibility scores, grades, overall rankings or invented assistant answers.
Manual AI-answer observations
Mentions and citations are separate lists. Each observation records the question, platform, date and response evidence. Unavailable answers are labeled.
Monthly monitoring
Reports changes in the evidence we actually collect. Limited observations cannot measure overall AI recommendation frequency.

What you can hold us to

  • Agreed, eligible improvements with authorized access.
  • Accurate business information based on facts you verify.
  • Documentation of changes and implementation verification.
  • Clear evidence and limits in the Signal Report.

What we won't promise

  • AI mentions, citations or recommendations.
  • Rankings, calls, leads or booked jobs.
  • A date when AI platforms will change their answers.
Free scan