Executive summary
State the collection date, market, providers, eligible coverage, headline findings, and the decision the report supports.
A client- and leadership-ready reporting structure that keeps the raw evidence, collection coverage, and limitations behind every conclusion.
A report, not a scorecard screenshot
Fill the template only after the collection is complete. Write the executive summary last so it reflects eligible results rather than the outcome you expected before testing.
Link every important statement to an answer, citation, prompt, or documented calculation. Put full transcripts and source URLs in an appendix so the main report stays readable without becoming opaque.
Six required sections
State the collection date, market, providers, eligible coverage, headline findings, and the decision the report supports.
Record the frozen prompts, provider surfaces, model identifiers, grounding requirements, locale, session state, and exclusions.
Report neutral discovery mentions separately from brand-named diagnostics, then show whether the brand led or merely appeared.
List claimed-domain citations, recurring third-party sources, unsupported claims, and source gaps by prompt.
Compare selected competitors at the answer level without presenting a small convenience sample as total market share.
Tie each action to stored evidence, name an owner, define completion evidence, and preserve the frozen set for the rerun.
Contextual next steps
Use the next resource that matches the decision at hand: define the program, inspect the method, or collect a comparable cross-provider baseline.
Run a frozen 25-question benchmark across four providers and inspect the underlying answer evidence.
02Review eligibility, denominators, coverage, citations, retention, and like-for-like rerun rules.
03Map cited pages to supported claims, brand effects, evidence gaps, owners, and actions.
Common questions
Include the test protocol, question set, provider coverage, brand mentions, prominence, citations, competitors, representation accuracy, raw evidence, limitations, and prioritized actions.
A composite can aid scanning, but it should never replace component metrics. Visibility, prominence, citations, accuracy, sentiment, and coverage answer different questions and should remain visible.
Update it after meaningful changes have had time to become retrievable. Use the same frozen prompts and protocol when the purpose is comparison; otherwise label the new collection as a different benchmark.
Yes. Replace the placeholders, keep raw evidence in an appendix, disclose the collection method, and avoid promising rankings, citations, or traffic from a directional snapshot.
Apply this resource
Preserve the questions, conditions, answers, citations, and failures so the next decision rests on inspectable evidence.