The short answer
To create an AI visibility audit report, define the brand, its competitors and the buyer prompts that matter, then run each prompt across ChatGPT, Perplexity, Claude, Gemini and Google AI Overviews and record who is mentioned. Score mention rate and share of voice, flag the high-intent prompts where the brand is missing, map each gap to a specific recommendation, and package it into a branded, client-ready report. A tool like Surva.ai automates the querying, scoring and reporting so the whole audit takes minutes.
Key takeaways
- Start with 15 to 50 high-intent buyer prompts and 3 to 5 named competitors.
- Run every prompt on ChatGPT, Perplexity, Claude, Gemini and Google AI Overviews, and keep the answers as evidence.
- Lead the report with mention rate and share of voice; the gaps are what sell the work.
- End with prioritized recommendations so the audit becomes your proposal.
On this page
The 6-step process
- 1
Define the brand, competitors and buyer prompts
Start with the brand, 3 to 5 real competitors, and the actual questions buyers ask AI: "best [category] tool," "[competitor] alternatives," "how do I [job]." Prioritize prompts with buying intent. Prompt tracking helps you find them.
For a local business, 15 to 25 prompts is usually enough; for a software company, 30 to 50 covering category, comparison and problem questions. Add a few brand prompts ("is [brand] good") to catch inaccurate claims AI makes about the client.
- 2
Run each prompt across the AI platforms
Query ChatGPT, Perplexity, Claude, Gemini and Google AI Overviews with each prompt and record which brands are mentioned and cited. Doing this by hand is tedious and inconsistent; AI visibility tracking runs every prompt across every platform automatically.
Keep the full answer text, the brands named, their order and any cited URLs. When the client asks "why does AI recommend them?", the stored answers are your evidence.
- 3
Score mention rate and share of voice
Calculate how often the brand appears across all prompts and platforms (mention rate) and how that compares to competitors (share of voice). These two numbers are the headline of the report and the thing clients remember.
Report both overall and per engine. A brand can be strong in ChatGPT and absent from Perplexity, and that split often points straight to the fix.
- 4
Identify the visibility gaps
Flag the high-intent prompts where competitors are cited and the brand is missing entirely. These gaps are the most persuasive part of the audit and the exact work you will propose. Competitor intelligence shows who is winning each one.
Sort gaps by intent first, then by how many engines miss the brand. A prompt like "best [category] for [their ideal customer]" where three competitors are named and the client is not is your headline finding.
- 5
Turn gaps into recommendations
An audit without next steps is just bad news. Map each gap to a concrete action: content to create, page structure or schema to fix, or authority signals to build. This turns the report into a proposal. Run an AI SEO audit to catch on-page issues too.
Group recommendations into three buckets clients understand: content (pages to create or rewrite), authority (third-party sites and reviews to target) and technical (crawler access, schema, rendering). Add rough effort for each so the plan feels realistic.
- 6
Package it into a branded report
Present the findings in a clean, client-ready format: executive summary, share-of-voice chart, prompt-by-prompt breakdown and prioritized recommendations. Agency Mode generates white-label reports you can hand straight to the client.
Keep the executive summary to one page. Most clients read the headline numbers and the gap list; the prompt-by-prompt appendix is there to back them up.
What to include in the report
Executive summary
The brand's overall mention rate and a one-line verdict on where they stand.
Share of voice vs competitors
A simple chart comparing the brand's AI appearances against named rivals.
Prompt-by-prompt breakdown
For each buyer prompt: is the brand mentioned, on which platforms, and who wins if not.
Prioritized recommendations
The specific actions to close the top gaps, which becomes your proposal.
Audit metrics, defined
Use the same definitions in every audit so results are comparable across clients and over time.
Swipe the table to see every column.
| Metric | How to calculate it | Why it matters |
|---|---|---|
| Mention rate | Prompts where the brand is named, divided by prompts checked | The headline number clients remember |
| Share of voice | The brand's mentions divided by all brand mentions in the answers | Shows position against named competitors |
| Average position | Where the brand appears in the list when it is named | First mentions carry more weight |
| Citation rate | Answers that link to the brand's own pages | Shows whether the client's content is used as a source |
| Engine coverage | Engines where the brand appears at least once | Reveals single-engine blind spots |