The AI Visibility Audit: 7 Things to Check Before You Spend Anything
A business can rank on Google's first page and still be completely invisible to ChatGPT, Perplexity, and Google AI Overviews. This 7-check audit tells you exactly where you stand, with a 0-21 point score, in under two hours.
A RankOS benchmark published in December 2025 found that 87% of U.S. businesses that rank on Google's first page do not appear in AI-generated answers (RankOS/GlobeNewswire, December 2025). The gap between traditional search visibility and AI visibility is not a minor technical nuance — it is the difference between being recommended and being ignored when a potential customer asks an AI assistant for help.
This guide gives you a structured, self-serve checklist with the exact prompts to run, clear criteria for passing and failing each check, and specific fixes for every gap you find. No paid tools required.
| Check | Method | Max Score |
|---|---|---|
| 1. Brand Recognition Test | Query each AI directly | 3 points |
| 2. Category Recommendation Test | Prompt each AI with category + city | 3 points |
| 3. GBP / Bing Places Completeness | Review your profile fields | 3 points |
| 4. Schema Markup Validation | Google Rich Results Test | 3 points |
| 5. AI Crawler Access | Inspect robots.txt | 3 points |
| 6. Review Health | Count and assess reviews | 3 points |
| 7. NAP Consistency | Cross-platform comparison | 3 points |
| Total | 21 points |
Score interpretation: 15-21 = AI Visible / 8-14 = Needs Work / 0-7 = Invisible
Check 1: Does Any AI Know Your Business Exists?
Ask each AI platform the exact query "What is [your business name]?" and evaluate whether it returns accurate, partial, or no information.
This is the most basic test of AI brand recognition. If a platform cannot identify your business when asked directly by name, it has no chance of recommending you when a customer asks a category question. The brand recognition test reveals whether your business exists in an AI platform's training data and live index — and whether that information is accurate.
Run this exact prompt on all five platforms:
"What is [Your Business Name]? Tell me about their services, location, and what they're known for."
Run this on: ChatGPT (chat.openai.com), Perplexity (perplexity.ai), Google AI (Google search with AI Overview enabled), Claude (claude.ai), and Gemini (gemini.google.com).
| Result | Score | What It Means |
|---|---|---|
| Accurate description, correct location, correct services on 4-5 platforms | 3 pts | Strong brand entity established |
| Partial info on 2-3 platforms, some inaccuracies | 1-2 pts | Entity exists but is incomplete or inconsistent |
| "I don't have information about this business" on most platforms | 0 pts | No established brand entity in AI knowledge bases |
If you fail: Your business has no established entity in the AI knowledge graph. Fixes include: submitting to Wikidata, getting your business mentioned in third-party articles and press coverage (brands with third-party media citations are 4x more likely to be referenced by AI), creating a detailed Google Business Profile, and ensuring your website's About page explicitly describes what your business does, where, and for whom.
Record your score: ___/3
Check 2: Do You Appear When Customers Ask for Recommendations?
Prompt each AI "Best [your category] in [your city]?" and check whether your business appears in the returned list.
This is the revenue-critical test. When a customer in your city opens ChatGPT and types "best accountant in Manchester" or "top plumber in Austin," they receive a short list of recommendations — typically 3-8 businesses. If you are not on that list, you do not exist for that customer. The category recommendation test is how you find out whether you are on the list.
Run these exact prompts on all five platforms:
"Best [your business category] in [your city]?"
"Top [your service type] near [your neighborhood or zip code]"
"Who should I call for [specific service] in [your city]?"
Document your results in a table:
| Platform | Appeared? | Position | Competitors Listed |
|---|---|---|---|
| ChatGPT | Yes / No | # | |
| Perplexity | Yes / No | # | |
| Google AI | Yes / No | # | |
| Claude | Yes / No | # | |
| Gemini | Yes / No | # |
| Result | Score |
|---|---|
| Appear on 4-5 platforms consistently | 3 pts |
| Appear on 2-3 platforms | 2 pts |
| Appear on 1 platform or occasionally | 1 pt |
| Appear on 0 platforms | 0 pts |
If you fail: The category recommendation test failure has multiple potential causes. ChatGPT sources recommendations primarily from Bing — so Bing Places completeness is critical. Perplexity favors freshness and third-party citations. Google AI and Gemini rely on your Google Business Profile and structured data. Claude indexes via Brave Search. A different fix may apply per platform. Note which platforms you appear on, as that tells you where your signal is already strong.
Record your score: ___/3
Check 3: Are Your Business Profiles Complete Enough to Feed AI?
Review your Google Business Profile and Bing Places listing field by field against the completeness checklist below.
Google Business Profile and Bing Places are primary data sources for AI recommendations about local businesses. ChatGPT answers local queries using Bing data, making Bing Places essential for ChatGPT visibility specifically. Google AI Overviews and Gemini pull directly from GBP. Research published in early 2026 found that cross-platform entity verification now accounts for over 70% of local AI ranking signals, and that AI enforcement of profile completeness is stricter than ever — businesses that had not posted an update or photo in 30+ days saw dramatic drops in AI impressions (Trebletree, 2026).
GBP Completeness Checklist
- Business name is exact legal name (no keyword stuffing)
- Primary category is specific, not generic (e.g., "Italian Restaurant" not "Restaurant")
- Secondary categories are set (up to 9 allowed)
- Business description is 750 characters, describes services and differentiators
- Services list is complete with individual service descriptions
- Hours are set including holidays
- At least 10 photos uploaded (exterior, interior, team, work samples)
- At least 1 GBP post within the last 30 days
- Questions & Answers section populated with common questions
- Website URL is correct and resolves
Bing Places Checklist
- Claimed and verified at bingplaces.com
- Name, address, phone match GBP exactly
- Categories selected
- Business description added
- Photos uploaded
| Result | Score |
|---|---|
| 12+ items checked on both platforms | 3 pts |
| 8-11 items checked | 2 pts |
| 5-7 items checked | 1 pt |
| Fewer than 5 or profiles unclaimed | 0 pts |
If you fail: Claim both profiles if unclaimed. GBP verification can take 1-14 days via postcard or video. Complete every field — partial profiles score significantly lower in AI recommendation systems. Set a monthly reminder to add a new GBP post and photo.
Record your score: ___/3
Check 4: Can AI Platforms Read Your Website's Structured Data?
Run your homepage and primary service pages through Google's Rich Results Test and confirm valid LocalBusiness JSON-LD schema is present with no errors.
Schema markup is the machine-readable layer that tells AI systems exactly who your business is, what you offer, where you are, and how to reach you. Without schema, AI platforms must infer this information from natural language text — which introduces errors and reduces confidence. Pages with valid structured data receive an average 43% higher rate of explicit AI mentions compared to pages without schema (wellows.com, 2025). The AI visibility boost from schema markup across multiple industry studies averages +36%.
How to Test
- 1 Go to search.google.com/test/rich-results
- 2 Enter your homepage URL
- 3 Enter your primary service page URLs
- 4 Check for: LocalBusiness detected / No errors / Required fields populated
Minimum LocalBusiness Schema Fields
{
"@type": "LocalBusiness",
"name": "Exact business name",
"address": {
"@type": "PostalAddress",
"streetAddress": "123 Main St",
"addressLocality": "Your City",
"addressRegion": "State/County",
"postalCode": "12345",
"addressCountry": "US"
},
"telephone": "+1-555-000-0000",
"url": "https://yourdomain.com",
"openingHoursSpecification": [...],
"geo": { "@type": "GeoCoordinates", "latitude": 0.000, "longitude": 0.000 },
"sameAs": ["https://g.co/your-GBP-link", "https://yelp.com/biz/your-listing"]
} | Result | Score |
|---|---|
| Valid LocalBusiness schema, zero errors, all required fields present | 3 pts |
| Schema present but with warnings or missing optional fields | 2 pts |
| Schema present with errors | 1 pt |
| No schema detected | 0 pts |
If you fail: Use Google's Structured Data Markup Helper (search "Google Structured Data Markup Helper") or TechnicalSEO.com's JSON-LD generator to create valid schema. Paste the output into your site's <head> section. Then retest with the Rich Results Test to confirm zero errors. Add FAQPage schema for any frequently asked questions visible on your pages.
Record your score: ___/3
Check 5: Can AI Crawlers Actually Access Your Site?
Navigate to yourdomain.com/robots.txt and inspect for blocks on the five major AI crawlers, then check your Cloudflare settings if applicable.
This is the most frequently missed technical failure in AI visibility. A business can have perfect content, complete profiles, and valid schema — and still be entirely invisible to AI systems because their robots.txt file or Cloudflare configuration blocks the crawlers that feed AI platforms. One documented case study describes a SaaS company that spent six months optimizing content with zero AI visibility gains, then discovered Cloudflare was blocking all AI bots by default. After allowing access, citations increased 217% within 30 days (Zelitho, 2025). In July 2025, Cloudflare changed its default configuration to block AI bots — meaning any site using Cloudflare that has not explicitly reviewed its bot settings may be blocking all AI crawlers without knowing it.
How to Check robots.txt
- 1 Open a browser and go to:
yourdomain.com/robots.txt - 2 Look for any of these user-agents followed by
Disallow: /
AI Crawlers to Check
| Platform | User-Agent(s) |
|---|---|
| OpenAI / ChatGPT | GPTBot |
| Anthropic / Claude | ClaudeBot, anthropic-ai, claude-web |
| Perplexity | PerplexityBot |
| Google (Gemini / AI Overviews) | Google-Extended |
| Microsoft / Copilot | Bingbot |
| Apple | Applebot-Extended |
Blocking entry (bad)
User-agent: GPTBot Disallow: /
Allowing entry (good)
User-agent: GPTBot Allow: /
If you use Cloudflare: Log into your Cloudflare dashboard, go to Security > Bots, and confirm you have not enabled "Block AI Scrapers" or similar rules introduced in the July 2025 update.
| Result | Score |
|---|---|
| All five major AI crawlers are explicitly allowed or not blocked | 3 pts |
| 1-2 crawlers blocked, others allowed | 1-2 pts |
| All AI crawlers blocked, or Cloudflare default block active | 0 pts |
If you fail: Edit your robots.txt to explicitly allow the AI crawlers you want to permit. Do not set a blanket Disallow: / under User-agent: * without specific Allow rules for AI bots. Review Cloudflare settings immediately.
Record your score: ___/3
Check 6: Do You Have the Review Health AI Systems Trust?
Count your reviews across all platforms, check their recency, and assess whether your response rate signals an active, trustworthy business.
Reviews are a primary trust signal for AI recommendation systems. When ChatGPT, Perplexity, or Google AI recommends a local business, it synthesizes review data from multiple platforms to assess credibility. Research published by SingleGrain in 2025 confirms that brands with verified and recent reviews receive 40% more mentions in AI-generated responses. The BrightLocal Local Consumer Review Survey (2026 edition) found that use of AI tools for local recommendations grew from 6% to 45% in a single year — and those AI tools heavily weight review recency and cross-platform presence. Businesses with fewer than 10 total reviews rarely appear in AI recommendation lists, while those with 50+ reviews across multiple platforms dominate in competitive categories.
Review Health Checklist
Run this audit across: Google, Bing, Facebook, Yelp, TripAdvisor (if relevant), and any industry-specific directories.
| Platform | Total Reviews | Avg Rating | Reviews in Last 90 Days | Owner Responses? |
|---|---|---|---|---|
| Yes / No | ||||
| Bing | Yes / No | |||
| Yes / No | ||||
| Yelp | Yes / No | |||
| Industry directory | Yes / No |
Scoring Criteria
| Result | Score |
|---|---|
| 25+ reviews on Google, reviews on 3+ other platforms, average rating 4.0+, reviews within 30 days, response rate 80%+ | 3 pts |
| 10-24 reviews on Google, some presence on 1-2 other platforms, average rating 3.5+ | 2 pts |
| 5-9 reviews on Google, limited other platform presence | 1 pt |
| Fewer than 5 reviews total or average rating below 3.5 | 0 pts |
If you fail: AI systems cannot confidently recommend businesses with sparse review histories. Create a systematic review request process: ask every satisfied customer at the end of the transaction with a direct link to your Google review page. Respond to every existing review — both positive and negative. Spread review generation across platforms, as Google-only concentration is a risk signal for AI systems that cross-reference multiple sources.
Record your score: ___/3
Check 7: Is Your Business Information Consistent Across Every Directory?
Search your business name across 10+ platforms and confirm that Name, Address, and Phone number are character-for-character identical everywhere.
NAP consistency — Name, Address, Phone — is the foundation of local entity verification. AI platforms do not recommend businesses they cannot reliably identify. When your name appears as "Smith Plumbing" on Google, "Smith Plumbing LLC" on Yelp, and "Smith & Sons Plumbing" on Facebook, AI systems face conflicting data about whether these are the same business. Research published by PlanDigi in 2025 found that 73% of customers lose trust in businesses that show different information online — and AI recommendation engines apply exactly the same verification logic before surfacing a recommendation. A 2025 case study of a regional law firm documented a 340% increase in AI-generated local search visibility after six months of standardizing NAP across 47 platforms (PlanDigi, 2025).
How to Check NAP Consistency
Search your exact business name in quotes on:
- Google ("Smith Plumbing" site:google.com)
- Bing Maps
- Apple Maps (search in Maps app)
- Yelp
- Yellow Pages
- BBB (Better Business Bureau)
- Foursquare
- Hotfrog
- Any industry-specific directories
For each listing, record what you find and compare against your canonical NAP:
| Platform | Name Listed | Address Listed | Phone Listed | Match? |
|---|---|---|---|---|
| Y / N | ||||
| Bing | Y / N | |||
| Apple Maps | Y / N | |||
| Yelp | Y / N | |||
| Y / N | ||||
| BBB | Y / N |
| Result | Score |
|---|---|
| NAP is identical across 8+ platforms with zero discrepancies | 3 pts |
| Minor discrepancies on 2-3 platforms (e.g., "St" vs "Street") | 2 pts |
| Significant discrepancies on multiple platforms | 1 pt |
| Major conflicts or business unclaimed on most platforms | 0 pts |
If you fail: Define one canonical NAP (use the exact format on your Google Business Profile as the master). Then systematically claim and update each platform. Start with the highest-authority directories: Google, Bing, Yelp, Facebook, Apple Maps. Use a service like BrightLocal or Whitespark for a citation audit if you operate across multiple locations. Businesses with 40+ accurate citations rank 53% higher in local AI search compared to those with fewer listings (fatjoe.com, 2025).
Record your score: ___/3
Your Total Score and What It Means
Add up your scores from all 7 checks.
| Total Score | Status | What to Do Next |
|---|---|---|
| 15-21 | AI Visible | You have the core signals in place. Focus on content quality, fresh reviews, and monitoring. |
| 8-14 | Needs Work | Prioritize the checks where you scored 0-1. Two or three targeted fixes can shift your visibility significantly. |
| 0-7 | Invisible | AI platforms cannot reliably find, verify, or recommend your business. Start with Check 5 (crawler access) and Check 7 (NAP), then work up. |
Most businesses score between 4 and 9 on their first audit. The checks that most commonly produce zero scores are: Check 5 (AI crawler access — often blocked by Cloudflare without the owner's knowledge), Check 4 (schema markup — most small business websites have none), and Check 7 (NAP consistency — accumulated discrepancies from years of listings created at different times with different formats).
What This Audit Does Not Cover
The 7-check audit above addresses the technical and structural foundations of AI visibility. It does not address content strategy, which is the second layer of AI visibility. Even with clean technical signals, a business with no content that AI systems can extract and cite will be limited to appearing in direct brand and category searches. The content layer includes: published articles and guides that AI systems cite as sources, structured FAQ content on your website, schema-tagged service descriptions, and topic cluster coverage that signals subject-matter authority.
The audit also does not cover off-page signals: mentions of your business in third-party articles, press coverage, and directory descriptions that AI systems pull from. Brands with third-party media citations are 4x more likely to be referenced by AI (RankOS, December 2025). A clean technical foundation is necessary — but it is not sufficient for maximum AI visibility.
How to Fix Your Lowest-Scoring Checks First
Prioritize repairs in this order, regardless of which checks failed. The impact-to-effort ratio is highest in this sequence:
- 1
Fix AI crawler access (Check 5) first.
If crawlers cannot reach your site, nothing else matters. This takes 15 minutes.
- 2
Fix NAP consistency (Check 7) second.
Conflicting entity data undermines every other signal. Start with your top 5 directories.
- 3
Complete your GBP and Bing Places (Check 3) third.
These are the primary data sources for the two most-used AI platforms (ChatGPT via Bing, Gemini via Google).
- 4
Add schema markup (Check 4) fourth.
A single implementation improves visibility across all platforms simultaneously.
- 5
Build review health (Check 6) as an ongoing process.
Reviews cannot be created overnight. Set up a system and commit to 30 days.
- 6
Track brand and category recognition (Checks 1 and 2) monthly.
These are outcome metrics — they improve as you fix the upstream signals.
Frequently Asked Questions
How long does an AI visibility audit take?
A manual audit using the 7 checks in this guide takes 45-90 minutes for a single-location business. Most time is spent on the category recommendation test (running prompts across five AI platforms) and the NAP consistency check (comparing your details across 10+ directories).
Do I need a paid tool to audit AI visibility?
No. All 7 checks in this guide can be completed manually for free using ChatGPT, Perplexity, Google AI, Claude, Gemini, Google's Rich Results Test, and direct inspection of your robots.txt file. Paid tools such as Otterly.ai, Semrush AI Visibility, and GEOReport.ai automate ongoing monitoring but are not required for an initial audit.
Why does my business appear on Google but not in AI answers?
A RankOS benchmark published in December 2025 found that 87% of U.S. businesses that rank on Google's first page are still invisible in AI-generated answers. Common reasons include blocked AI crawlers in robots.txt, missing schema markup, insufficient third-party review coverage, and inconsistent NAP data across directories.
Which AI platforms should I test for local business visibility?
The five most important platforms for local businesses are ChatGPT (indexes via Bing), Perplexity (independent crawler with freshness emphasis), Google AI Overviews (Google index), Claude (indexes via Brave Search), and Gemini (Google index and Knowledge Graph). Each platform has different citation signals, so a business can rank highly on one while being invisible on another.
What is a good AI visibility score on this audit?
Scores of 15-21 out of 21 indicate AI Visible status — your business has the foundational signals AI platforms need to recommend you. Scores of 8-14 mean your business Needs Work; you are partially visible but missing critical signals. Scores of 0-7 mean Invisible; AI platforms cannot reliably find, verify, or recommend your business.
How do I check if AI bots are blocked on my site?
Navigate to yourdomain.com/robots.txt in a browser. Look for Disallow rules under User-agent: GPTBot, User-agent: ClaudeBot, User-agent: PerplexityBot, or User-agent: Google-Extended. Any Disallow: / entry under these agents blocks that platform entirely. Also check your Cloudflare dashboard under Security > Bots, as Cloudflare began defaulting to AI bot blocking in July 2025.
How many reviews does a business need to appear in AI recommendations?
Research published by BrightLocal in 2026 and SingleGrain in 2025 indicates that businesses with fewer than 10 reviews rarely appear in AI recommendation lists. Top-ranking businesses in competitive local categories typically have 50+ reviews across multiple platforms, with new reviews appearing monthly. Brands with verified and recent reviews receive 40% more mentions in AI-generated responses.
What schema markup does a local business need for AI visibility?
The minimum requirement is LocalBusiness JSON-LD schema including name, address (PostalAddress), telephone, openingHoursSpecification, geo coordinates, url, and sameAs links to your Google Business Profile and major directories. Adding FAQPage schema for common questions and adding an llms.txt file at your domain root provides additional signals to AI crawlers.
How often should I repeat this AI visibility audit?
Run a full 7-check audit every 90 days at minimum. AI platforms update their indexes and ranking signals continuously. The category recommendation test (Check 2) should be run monthly, as competitive position in AI recommendations shifts faster than traditional search rankings.
What is NAP consistency and why does it affect AI search?
NAP stands for Name, Address, and Phone number. AI platforms cross-reference your business data across multiple directories and platforms before recommending you. When your name appears differently across platforms, AI systems cannot confidently verify your identity. Research from 2025 shows that 73% of customers lose trust in businesses that show different information online, and AI recommendation engines apply the same verification logic.
Key Takeaways
- 1
87% of businesses that rank on Google are still invisible in AI-generated answers, making this audit relevant for almost every local business (RankOS, December 2025).
- 2
The 7 checks cover the foundational signals AI platforms use: brand entity recognition, category presence, structured profiles, schema markup, crawler access, review trust, and NAP consistency.
- 3
Scores of 15-21 indicate AI Visible status; 8-14 indicates Needs Work; 0-7 indicates Invisible.
- 4
AI crawler blocking — often caused by default Cloudflare settings changed in July 2025 — is the most impactful single fix because it prevents all other signals from being read.
- 5
NAP consistency across 8+ directories is the difference between AI platforms confidently identifying your business and silently skipping it.
- 6
Reviews on multiple platforms, not just Google, are required for AI systems to recommend you with confidence.
- 7
Schema markup (LocalBusiness JSON-LD) remains underused — most small business websites have none — and provides a direct, machine-readable path for AI extraction.
Published February 22, 2026. Research current as of February 2026. Sources: RankOS/GlobeNewswire (December 2025), BrightLocal Local Consumer Review Survey (2026), SingleGrain AI Reviews Analysis (2025), PlanDigi NAP Consistency Study (2025), Zelitho GEO Readiness Checklist (2025), Trebletree GBP Local Visibility Playbook (2026), Cloudflare AI Bot Blocking Announcement (July 2025), wellows.com AI Visibility Audit Checklist (2025).