AI Visibility Audit in 30 Minutes: A Step-by-Step Framework

What Is an AI Visibility Audit?

An AI visibility audit is a structured review of how AI language models — ChatGPT, Google Gemini, Perplexity, and Claude — perceive, understand, and recommend your business. Unlike a traditional SEO audit that focuses on keyword rankings and technical page health, an AI visibility audit examines entity signals, content authority, and third-party validation: the three inputs AI models use to decide which businesses to surface in conversational recommendations.

The audit answers five core questions:

• Does ChatGPT (and other AI engines) know your business exists?
• Does it understand what your business does and who it serves?
• Does it have enough trust signals to recommend you confidently?
• Are there specific queries where competitors appear and you don't?
• What is the highest-ROI fix you can make this week?

You do not need to run this audit every day. A quarterly audit with monthly spot-checks is sufficient for most businesses. Use this first run to establish your baseline — every subsequent audit measures improvement against it.

What You Need Before You Start

Open these browser tabs before beginning. The audit is easier when you're not switching between tasks:

Also prepare three things before running queries: (1) your exact business name as you want it recognized, (2) three to five keywords describing what your business does, and (3) the geographic area you primarily serve, if applicable. These feed directly into your query set in Step 1.

Step 1: Run a Live AI Visibility Test (10 minutes)

This is the most direct signal you can gather. You're going to ask the AI platforms the same questions your customers ask, and record what comes back.

Build your query set

Write out five query types, customized to your business. Use this template:

Run the queries and record results

Open a fresh conversation (no history) in each AI platform and run all five queries. In your blank document, record for each query and platform:

• Is your business mentioned at all? (Yes / No)
• If yes: is it a strong recommendation, a passing mention, or a neutral reference?
• Which competitors are mentioned instead of (or alongside) you?
• Does the AI describe your business accurately?

Note: ChatGPT without web browsing enabled draws purely from training data. Perplexity always cites live web sources. Both matter — training data visibility is harder to build and more durable; web citation visibility is faster to influence.

After running all queries, you should have a clear picture: are you invisible, partially visible, or well-represented? Most businesses fall into one of three categories:

Step 2: Audit Your Entity Signals (7 minutes)

Entity signals are the structured and semi-structured data points that help AI models identify your business as a known, trustworthy entity. A weak entity profile is the most common reason businesses are invisible in AI search even when they have strong Google rankings.

Schema markup check (3 minutes)

Go to Google's Rich Results Test and enter your homepage URL. Look for:

• Organization schema: Does it show your business name, URL, logo, and sameAs links? Missing Organization schema is the most common gap.
• LocalBusiness schema: If you serve a specific geography, do you have address, phone, and opening hours structured?
• sameAs property: This is critical. The sameAs field should link to your LinkedIn, Crunchbase, Wikipedia (if applicable), and major directory profiles. It directly tells AI models which external entities are the same as your business.

If you have no schema at all, note this as a Priority 1 fix. If schema exists but lacks sameAs links, that's a Priority 2 fix. Both can be completed in under an hour by a developer. For a deeper dive on exactly what to implement, see our schema markup guide for AI search.

Directory consistency check (4 minutes)

Open a new tab and search for your business name on each of these platforms. Record whether your listing exists and whether the name, description, and category are consistent:

• Google Business Profile
• LinkedIn Company Page
• Crunchbase
• Yelp (if relevant)
• One or two industry-specific directories for your sector

Inconsistency — different business names, slightly different descriptions, wrong categories — fragments your entity profile. AI models reconcile data across sources; conflicting signals reduce confidence and lower recommendation likelihood. The fix is simple: standardize your name, tagline, and service description across all platforms.

Step 3: Content Gap Analysis (5 minutes)

AI models learn your expertise from what they find written about your topics. If you have not published content on the questions your customers ask, you have zero topical signal to draw on. This step identifies the most important content gaps fast.

Map the top 10 questions to existing content

List the 10 most common questions your sales team or customer support hears. For each question, check whether you have a dedicated page (article, FAQ, guide, or landing page) that answers it directly. Not just mentions the topic in passing — actually answers it as the primary subject of a page.

A question without a dedicated page is a content gap. Prioritize gaps where:

• The question is frequently asked by high-value prospects
• Competitors have dedicated content but you don't
• The question appeared in your Step 1 query set (meaning AI platforms are actively answering it for people)

Check for "best X" and comparison coverage

Run this quick search in Google: site:yourdomain.com "best" OR "compare" OR "vs". If you have zero pages with comparison or list-style content, that's a significant gap. Listicles and comparison guides are among the most-cited content types in AI recommendations — the model needs a reference to pull your name from when answering "what's the best [X]" queries.

Step 4: Third-Party Validation Audit (5 minutes)

Third-party validation is independent evidence that your business is credible and delivers results. AI models weight this heavily because it is hard to fake at scale. A business mentioned positively across review platforms, industry publications, and editorial roundups is far more likely to be recommended than one whose only digital footprint is its own website.

Review platform check (2 minutes)

Check your current review status on:

• Google Reviews — count and average rating
• Trustpilot — do you have a profile? What's your score?
• G2 or Capterra (for software/services businesses)
• Industry-specific review sites for your sector

Record the total number of reviews and your average rating across platforms. Under 25 total reviews across platforms is a red flag for AI visibility. Between 25 and 100 is functional but weak. Over 100 with a 4.5+ average is a strong validation signal.

Press and editorial mentions (3 minutes)

Search Google for: "[Your Business Name]" -site:yourdomain.com. This shows mentions of your brand on sites other than your own. What you want to see:

• Mentions in trade publications or industry blogs
• Inclusions in "best of" or comparison roundups by third-party authors
• Quotes or contributed articles in relevant media
• Podcast appearances or event listings

If results are sparse — just your own social profiles and a few directory listings — you have a third-party validation gap. This is typically the hardest thing to fix quickly, which is why the businesses that have strong press footprints enjoy a durable competitive advantage in AI search.

Step 5: Score and Prioritize (3 minutes)

Use this scoring table to assess your current AI visibility across the five dimensions of the audit. Score each area 1–3 based on what you found:

Add your scores. The maximum is 15. Use this to interpret your result:

• 5–7: Foundational stage. Your business has minimal AI presence. Start with entity building (schema + directories) — it's the fastest, highest-leverage fix.
• 8–10: Developing stage. You have some presence but significant gaps. Focus on content authority and review generation in parallel.
• 11–13: Established stage. Solid foundation. Focus on expanding your content depth and acquiring more editorial mentions to push into stronger recommendation territory.
• 14–15: Advanced stage. You're well-positioned. Focus on monitoring for accuracy issues and defending against competitors who are catching up.

What to Do After the Audit

The most common mistake after an AI visibility audit is trying to fix everything at once. The areas with the highest ROI are entity signals (schema + directories) and content gaps, in that order. Third-party validation has the biggest long-term impact but takes the longest to build — start it now as a background program, but don't let it block your first-week fixes.

Here is a practical prioritization order for most businesses coming out of their first audit:

The most important thing is to retest. Run the exact same five queries from Step 1 again in 30 days. Changes to entity signals tend to show up in Perplexity (which crawls the web in real-time) faster than in ChatGPT (which uses training data), so use Perplexity as your early signal and ChatGPT as your longer-term validation. For a more detailed tactical program, the 90-day ChatGPT visibility guide covers the full execution plan step by step.

The Fast Path: Use the Free AI Visibility Score Tool

If you want a baseline score before running the manual audit — or want to sanity-check your findings — Magna's free AI Visibility Score tool runs a standardized test across ChatGPT, Perplexity, and Gemini in under 60 seconds. It checks five dimensions of AI presence and returns a score with specific findings. It's the fastest way to understand your starting position before investing time in the manual steps above.

After you have your score and manual audit findings, you'll have everything you need to build a prioritized AEO action plan. Understanding what AEO is and how it works gives you the strategic context behind each fix — worth reading alongside this audit if you haven't already.