An AI search visibility audit for a SaaS is a structured pass over your buyers' real questions, run through the major AI engines, logging whether you are cited, how you are described, and which sources drive the answers. It takes one to two days and produces a baseline plus a prioritized fix list.
Build the question set first
The audit is only as good as the questions. Pull 20-30 from the places buyers actually ask: sales call objections, support tickets, comparison and pricing queries, and category questions like 'best tool for X'. Cover every funnel stage, from 'what is [category]' to '[you] vs [alternative]' to 'is [you] worth it'.
What to log per engine
| Field | What to record | Why it matters |
|---|---|---|
| Cited? | Yes / no / competitor only | Your inclusion rate is the headline metric |
| Description | Accurate / outdated / wrong | Misdescription costs deals silently |
| Sources | Domains the answer draws on | These are the pages you must win |
| Sentiment | Positive / neutral / negative | Tone shapes shortlists |
| Competitors | Who else is named | Feeds share-of-voice math |
From findings to fix list
- 1Group gaps: not cited, misdescribed, out-cited by a competitor.
- 2For each, identify the source pages driving the answer.
- 3Rank fixes by impact (question value) times effort (own page edit vs third-party outreach).
- 4Ship the top five, then re-run the audit in 30 days against the same question set.
