Direct Answer The AI Visibility Audit is a repeatable method for measuring a consumer brand's presence inside AI engines. It runs in four steps: build a prompt set, test across all major engines, record three data points per prompt, and date the results as a baseline. It is the first action of any GEO program and the standard method for measuring Share of Model.
The AI Visibility Audit — four steps
Step | Action | Output |
1 | Build a 15–25 prompt set from real shopper questions | The Prompt Map |
2 | Run every prompt through ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews | Raw answers |
3 | Record: brand appears? how described? sources cited? | Three-point dataset |
4 | Date and save the results | Quarterly baseline |
Step 1 — Build the prompt set. Write the questions a real shopper asks, not keywords. Aim for 15–25 across the category. This set becomes the brand's Prompt Map.
Step 2 — Test every engine. Run each prompt through ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews. Coverage varies by engine; test all.
Step 3 — Record three data points. For each prompt: does the brand appear, how is it described, which sources did the engine cite. The cited sources are the brand's GEO target list.
Step 4 — Date the baseline. The audit's value is longitudinal. Re-run quarterly; the movement is the truest measure of GEO progress.
The three diagnostic outcomes The audit surfaces one of three states: the brand is absent (competitors own the answer), present but mis-described (an entity-data problem — see The Citation Stack), or present and accurate (the job becomes defense).
Prompt examples
"best [category] for [need]"
"what [product] should I buy"
"is [brand] any good"
"[brand] vs [competitor]"
"most trusted [category] brands"
FAQ
What is the AI Visibility Audit?
The AI Visibility Audit is a four-step method used to measure a brand’s presence across major AI engines and establish a baseline for visibility performance.
How often should the audit run?
Quarterly. Movement between audit periods helps measure progress and track the impact of a GEO program over time.
What does the audit produce?
The audit typically produces:
A baseline Share of Model (SOM) figure
A Prompt Map
A GEO target list of cited sources
Related: Share of Model · From Keywords to Prompts · GEO for Consumer Brands
