Measurement methodology

Every number should answer a business question.

Scope keeps metrics simple in the report and preserves the numerator, denominator, evidence, model, market, date, and eligibility rules underneath.

The direct answer

Scope measures recommendation behaviour from dated AI answers and does not infer visibility from rankings, crawler access, or one unexplained score.

01

How often are you recommended?

Recommendation rate is the percentage of eligible unbranded buyer answers that actively recommend the monitored domain. Recognition or a passing mention does not count as a recommendation.

  • Numerator: answers that recommend or prefer the domain.
  • Denominator: completed, relevant, eligible buyer answers.
  • Branded controls and failed answers are excluded.

02

How often are you part of the discussion?

AI visibility records eligible answers where the domain is recommended, preferred, mentioned, cautioned, or rejected. It describes presence, while recommendation rate describes active selection.

03

Who wins the same buyer needs?

Competition is measured inside the same questions, assistants, market, and collection window. A business must recur across enough independent evidence before Scope presents it as a meaningful competitor.

04

What evidence shapes the answer?

Scope records cited pages and groups them as owned, competitor, or independent sources. Search infrastructure and unusable URLs are excluded from source conclusions.

05

What counts as meaningful movement?

Scope compares compatible audits only. A change becomes convincing when it recurs across questions, assistants, or collection periods instead of appearing in one favourable response.

06

What the result cannot prove

An audit is a dated sample of generated answers. Scope cannot guarantee future recommendations, establish exact prompt demand without external panel data, or claim that one website change caused movement without sufficient repeated evidence.

Last reviewed 2026-09-17