Free 10-minute AI Search Diagnostic Request yours ↗

OBSERVED OUTCOMES, EXPLAINED HONESTLY

Know what changed—and what the numbers mean.

Separate the work completed, the answers observed, and the commercial outcomes you can actually measure.

01 / OBSERVED OUTCOMES, EXPLAINED HONESTLY

Use definitions that make comparisons possible

Report the unit of measurement and the denominator. Changing a question set can change a percentage even when the underlying brand presence has not improved.

  • Answer inclusion: sampled answers containing the brand ÷ sampled answers reviewed.
  • Mentions: occurrences of the brand name; repeated names in one answer are not multiple included answers.
  • Citation presence: sampled answers with a linked brand-owned URL ÷ sampled answers reviewed. Report third-party citations about the brand separately.
  • Accuracy: verified claims judged correct ÷ claims evaluated, using agreed business facts. Disclose partial or unassessable claims.
  • Competitive visibility: brand appearances or mention share within a defined sample and competitor set. State the chosen formula.
  • Referral traffic: identifiable visits from a source where tracking allows. Missing referrers can limit coverage.
  • Qualified inquiry: an inquiry meeting an agreed business criterion, evaluated separately from visibility.
02 / OBSERVED OUTCOMES, EXPLAINED HONESTLY

Illustrative reporting example—not actual results

Fictional sample: 20 questions, two platforms, one observation per question per platform, producing 40 answers. Example market: a fictional local service area. Observation dates: not applicable; this example was not run.

  • 8 of 40 answers include the fictional brand: 20% answer inclusion.
  • 4 of 40 answers link to a brand-owned page: 10% citation presence.
  • 9 of 12 evaluated factual claims match the fictional approved facts: 75% accuracy in that claim sample.
  • These are explanatory numbers, not a benchmark, real audit, trend, or client result.
  • A real report includes exact observation dates, platform context, geography, prompt wording, and evidence access.
03 / OBSERVED OUTCOMES, EXPLAINED HONESTLY

Keep the change log beside the metrics

An implementation record should say what was published, the relevant URL, approved scope, verification, and the observation questions it relates to.

  • Illustrative change: clarify roof replacement service scope on the fictional service page.
  • Status: example of a proposed action—not work completed for a client.
  • Verification in a real engagement: inspect published text, links, access, and relevant markup.
  • Retest: repeat the related category and branded questions; report all observations.
04 / OBSERVED OUTCOMES, EXPLAINED HONESTLY

Connect visibility with engagement carefully

Review relevant referral visits, inquiry paths, form submissions, qualification, and customer-reported discovery sources where configured. Treat those as different evidence types. A customer mentioning AI is useful context but not complete attribution.

  • Do not infer revenue from a brand mention or apply an invented conversion rate.
  • A form submission is not automatically a qualified inquiry.
  • A diagnostic request is not a confirmed appointment.
  • Missing tracking and platform variability must remain visible in the interpretation.
05 / OBSERVED OUTCOMES, EXPLAINED HONESTLY

Agree the reporting scope

Define cadence, questions, platforms, markets, evidence retention, report access, and business indicators before recurring work begins. Adding a new question group should be disclosed rather than blended into a historical comparison.

See the methodology ↗

YOUR NEXT CHAPTER OF DISCOVERY

Find the gaps. Understand the opportunities.
Build your next move.

Take a focused look at your AI search presence and discuss where a stronger answer could begin.

Free. No purchase required. A focused diagnostic—not a full audit.