What Is a Brand Prompt Audit and How Do I Run One?

A brand prompt audit is a structured test of how AI engines answer the specific questions your buyers ask, measuring whether your brand appears, in what position, with what sentiment, and against which competitors. You assemble a representative prompt set across buyer intents, run each prompt on every engine you care about, and score the results into a repeatable spreadsheet.

What does the audit measure?

A complete audit is 360-degree: it spans buyer intents and engines rather than one query on one tool. Score each prompt on:

• Presence, does your brand appear at all in the answer?
• Position, first mention, mid-list, or last?
• Sentiment, described positively, neutrally, or with a caveat?
• Competitors, which rivals appear, and how often?
• Citations, which of your URLs (if any) are linked as sources?

Recording all five turns anecdotes into a scorecard you can trend over time.

How do I build the prompt set?

Group prompts by intent so the audit reflects a real buyer journey. Problem-aware prompts ("how do I reduce SaaS onboarding drop-off"), solution-aware prompts ("best onboarding software for B2B"), and brand-aware prompts ("is [your brand] good for enterprise") each surface different visibility gaps. Mine your own site search, sales-call notes, and support tickets for the exact phrasing buyers use. Test on the engines your audience uses, ChatGPT, Perplexity, Claude, Gemini, and Google AI Overviews cover most B2B research.

How do I score it into a spreadsheet?

Build a grid with prompts as rows and engines as columns, then log presence, position, and sentiment per cell. Assign a simple weight, say 3 points for a first mention, 1 for a later mention, 0 for absence, and sum per engine to get a share-of-voice snapshot. Because AI answers vary between sessions, run each prompt in a fresh, logged-out session two or three times and record the mode.

Manual auditing works for a first pass but degrades fast past a few dozen prompts and engines. Prompt research automates the collection so your audit becomes a living tracking baseline instead of a one-time spreadsheet, letting you re-run the exact set after each content change and attribute movement to what you shipped.