AI Visibility Audit: Step-by-Step Guide
An AI visibility audit reveals how AI search engines see your brand — and why they might recommend competitors over you. Here's how to do it.
Most brands have no idea what AI assistants actually say about them. You can check traffic, rankings, and social mentions — but the answers ChatGPT and Perplexity give about your category are a blind spot. This guide walks you through a repeatable audit with the questions, scoring, and action plan that turn findings into fixes.
What is an AI Visibility Audit?
An AI visibility audit is a systematic analysis of:
- How often AI mentions your brand
- Which sources AI cites for your category
- Why competitors are recommended over you
- What technical gaps prevent AI from understanding your brand
The audit treats AI assistants as a channel with its own measurement. The output isn't a single number — it's a map of where you're visible, where you're absent, and which pages or gaps explain the difference.
How to Run an AI Visibility Audit
Step 1: Define Your Questions
List 10-15 buying questions your customers ask:
- "Best [your category] tools"
- "[Your brand] alternatives"
- "[Your brand] vs [competitor]"
- "Is [your brand] safe?"
- "[Your brand] pricing"
Aim for a mix of category questions ("best tools for X"), comparison questions ("A vs B"), and trust questions ("is A legit"). Category questions measure general visibility; comparison and trust questions measure whether you win the moments that drive purchases. If you're unsure what matters most, mine your sales inbox, support tickets, and Google autocomplete.
Step 2: Test Each Question
Ask each question to:
- ChatGPT (GPT-4o)
- Perplexity
- Google AI Overviews
Record:
- Is your brand mentioned?
- Is your brand recommended?
- What sources does AI cite?
Run the same question set across all three engines in one sitting so results are comparable, and log the exact wording and date — AI outputs change frequently, so you can compare like with like later. Copy the full answer text, not just whether you were mentioned; the surrounding sentences show why AI prefers a competitor.
Step 3: Analyze Results
Calculate your scores:
- Mention Rate = Times mentioned / Total questions
- Recommendation Rate = Times recommended / Total questions
- Citation Rate = Times cited / Total questions
These three rates tell different stories. High mentions with low recommendations means AI knows you exist but doesn't trust you as the answer — usually a content-depth problem. Low mentions point to retrieval problems: weak structure, missing entity signals, or thin coverage of buyer questions. Also note which sources AI cites most often — those are your real competitors.
Step 4: Identify Gaps
For each question where you're NOT recommended:
- Who IS recommended?
- What sources does AI cite?
- What do they have that you don't?
Be specific here. If AI cites a competitor's G2 profile, you need review presence, not more blog posts; if it cites their comparison page, you need one of your own; if it cites Reddit, you need community presence. Naming the asset that beat you turns "we need more authority" into a concrete task list.
What an AI Visibility Audit Reveals
Technical Gaps
- Missing Schema markup
- robots.txt blocking AI crawlers
- Poor page structure
- Slow load times
Technical gaps are the cheapest to fix and the most commonly missed. A single robots.txt rule blocking an AI crawler can make everything else invisible. Check your logs for AI user agents, validate your structured data, and keep key pages fast. These fixes rarely take more than a few days, and they compound with everything else.
Content Gaps
- No comparison pages
- No FAQ pages
- No structured pricing
- Thin content on key pages
Content gaps show up when AI finds you for category queries but not for decision queries. If your pricing is buried and you have no "vs" pages, the model has nothing clean to cite at the moment of purchase. Comparison pages and structured FAQs are the highest-ROI additions — see our guide to AI search optimization for the full playbook.
Authority Gaps
- No G2/Capterra presence
- No Reddit discussions
- No editorial coverage
- Weak backlink profile
Authority gaps explain why AI answers "best tools" with your competitors even when your content is better. AI systems weight third-party signals heavily, and a competitor with dozens of G2 reviews and active Reddit discussions wins the citation almost every time. Closing this gap takes months — so start it first.
Free AI Visibility Audit
Run a free audit with Gerush:
- Enter your domain
- System runs 15 questions × 3 AI models
- Get your scores in 2 minutes
- See exactly where you're losing to competitors
Doing the manual audit once teaches you the mechanics; automating it turns audit into monitoring. If you'd rather skip the spreadsheets, use our free AI visibility checker for an instant baseline, and keep the manual deep-dive for investigating a specific question or competitor.
After the Audit: Action Plan
Priority 1: Technical Fixes (Week 1)
- Add Schema markup
- Fix robots.txt
- Create llms.txt
Priority 2: Content Creation (Week 2-3)
- Create comparison pages
- Build FAQ pages
- Add structured pricing
Priority 3: Authority Building (Week 4-6)
- Get G2 reviews
- Participate in Reddit
- Submit to editorial lists
Priority 4: Monitoring (Ongoing)
- Track weekly changes
- Re-test monthly
- Adjust strategy
The order matters: technical fixes unblock everything else, content gives AI something to cite while authority accumulates, and monitoring turns this from a one-time project into a durable advantage. AI models update constantly — the brand that re-tests monthly sees shifts months before the brand that checks once a year.
How Often Should You Audit?
Run a full audit quarterly and a light check monthly — as small as re-testing your five key questions across two engines. Anything more frequent is noise, and anything less frequent means you'll miss a competitor's new page or a model update reshuffling the answers.
Common Mistakes
The most common mistakes are testing too few questions, only asking category queries, and quitting after the first run. A single "best tools" question tells you almost nothing; ten to fifteen questions across categories is where patterns emerge. Treat the first audit as a baseline, not a verdict — the value is in the second run, which shows whether your fixes work.
Start Your Audit Today
The first step to improving AI visibility is knowing where you stand.
Run a free AI visibility audit now and get your baseline score.