Reusable audit report outline

This template turns scattered AI search checks into a structured report. Use it after running a stable prompt set across answer engines. The finished report should explain what was tested, what sources were cited, where competitors appeared, and which page updates should happen next.

Executive summary block

Start with a short summary that a founder, SEO lead, or product marketer can understand without reading the full prompt log. State the audited brand, category, date range, engines tested, strongest signal, weakest signal, and the highest-priority fix.

  • Brand and category audited.
  • Engines tested and number of prompts per engine.
  • Official-site citation rate versus third-party citation rate.
  • Top competitor names that appeared repeatedly.
  • One recommended page update for the next 14 days.

Prompt evidence matrix

Use a matrix to avoid cherry-picking one answer. The goal is to show patterns across prompt types and engines. If one engine cites the official site and another cites only review pages, the report should preserve that difference.

Prompt typeEvidence to recordDecision it supports
Category promptMentioned brands, cited overview URLs, missing official pagesWhether the product is understood as part of the category
Alternatives promptCompetitors, comparison sources, wording of buyer criteriaWhether an alternatives page or proof page is needed
Pricing promptPricing pages cited and answer accuracyWhether pricing copy and schema need updates
Integration promptDocs, marketplace pages, and setup tutorials citedWhether public integration proof is strong enough

Source review checklist

After collecting prompt results, review the cited URLs before changing copy. The report should identify whether answer engines prefer official pages, review sites, directories, tutorials, forums, or competitor docs.

  • Group cited URLs by source type and owner.
  • Mark whether each cited page answers the prompt better than the official site.
  • Flag outdated or inaccurate answers separately from missing citations.
  • Connect every recommended fix to a specific URL and prompt type.
  • Set a retest date using the same prompt wording.

Action plan format

Close the report with a short action plan grouped by crawlability, entity clarity, answer snippets, citation proof, and comparison coverage. Each action should include an owner, expected output, source requirement, and retest condition so the report creates work rather than another dashboard.

Fill-in example

A first audit report for a fictional product like AcmeFlow could say: the brand appeared in two category prompts, was absent from alternatives prompts, and was cited only through a launch directory rather than the official product page. The immediate recommendation would be to improve crawler access, add a clearer answer block, and create one source-backed comparison brief before retesting.

Review handoff

Send the report with the exact prompt set, answer engine names, run date, cited URLs, competitor mentions, screenshots if needed, and a list of recommendations that were rejected during review. The next reviewer should be able to rerun the same prompts, compare citation changes, and understand why each action was prioritized.

Common failure modes

Weak audit reports treat one answer as a ranking, mix prompts between test runs, or recommend new content without checking whether official pages are crawlable. Another failure is ignoring third-party sources that answer the query better than the SaaS site. The report should explain those gaps instead of turning every finding into a writing task.

Generate a SaaS GEO audit report