AI Visibility Assessment / GEO Audit
A GEO audit tests the real recommendation questions your customers ask across the major AI platforms, documents whether your business is named and accurately described, identifies the competitors named instead, examines the sources being cited, and turns those findings into a prioritized implementation plan. $5,000, one flat fee, about 10 business days.
By JoLyn Laney, FounderLast updated: October 3, 2026What is a GEO Audit?
A GEO audit (Generative Engine Optimization audit), also called an AI Visibility Assessment, is a structured test of how AI platforms such as ChatGPT, Perplexity, Gemini, Microsoft Copilot and Google AI Mode respond to the questions your customers ask. It records whether your business is named, whether it is recommended, whether the facts are right, which competitors appear instead, and which sources are cited, then delivers a prioritized plan. It measures and recommends. It does not promise a recommendation.
New to GEO? Read the full definition and the five signals AI tools readWhy Measure Before You Change Anything
Avante tested how AI recommends local businesses across 60 markets. The findings are the reason this service starts with measurement and not with a list of tactics.
12.0% reached the first five. Of the businesses the AI itself listed as operating in a market, 12.0% appeared in its first five recommendations.
Half the sources were third-party. Of 1,128 citations, 49.6% pointed to individual business websites and 49.1% to third-party sources such as directories, review platforms, publishers, and community threads.
The leading source changed by industry. A legal directory led for personal injury law, a real-estate portal for agents, an insurance directory for dentists, and Reddit for restaurants and HVAC.
Source: Avante's AI Visibility Index, July 2026 edition. 120 live queries on Perplexity, 6 industries, 10 US metros, run July 6, 2026. One platform, one day. Full limitations are on the research page.
A general study cannot tell you which sources matter in your market. The audit runs the same kind of test on your questions, your competitors, and your city.
What We Test
Every finding in the report follows the same path, from a customer's question to a specific action.
- 1
Customer question
A real buying question from your query panel, asked the way a customer would ask it.
- 2
AI answer
The full response from each platform, recorded with the platform, model, and date.
- 3
Brand and competitor recommendations
Every business named, in order, with your business marked as named, recommended, or absent.
- 4
Cited sources
The websites the platform showed with the answer, grouped by type.
- 5
Likely gaps
Where your business is missing, thin, or inconsistent in the sources that appeared.
- 6
Recommended action
A specific, prioritized ticket. Not a general tip.
This follows the six levels of the Avante AI Visibility Measurement Framework: presence, recommendation, accuracy, citation and source context, competitive share, and commercial outcome.
What You Receive
Twelve deliverables, each one a document or a session you can point to.
Buying-Intent Query Panel
A fixed list of the questions your customers actually ask when they are choosing a business like yours, weighted toward questions that do not include your name. The panel is documented in the report and reused for every re-measure.
Multi-Platform Baseline Testing
The panel is run on ChatGPT, Perplexity, Gemini, Microsoft Copilot, Google AI Mode and Grok, in the real apps customers use wherever that testing is available (Grok through its developer version). Each question is run multiple times and reported with a confidence range, because AI answers vary and single-run tests mislead.
Brand Mention and Recommendation Findings
For every question and platform: whether your business was named, whether it was actually recommended, and in what position. Mentions and recommendations are reported separately.
Competitor-Share Findings
Who is named when you are not: your share of the answers in your market next to the businesses and directories that hold the rest.
Citation and Source Analysis
The sources each platform showed with its answers, grouped by type, so you can see which websites, directories, review platforms, and publishers sit behind the recommendations in your market.
AI Accuracy Findings
What AI gets wrong about your business. Every factual error, omission, and outdated claim is documented against a fact list you have confirmed.
Technical, Entity, and Content Evaluation
Structured-data validation, entity consistency, content structure, and verification that GPTBot, ClaudeBot, PerplexityBot, and Google's crawlers can actually reach your pages, including CDN and firewall blocks.
The Avante Scorecard
18 on-site readiness factors plus 9 off-site authority signals, each scored with evidence and weighted for your industry. Every report passes an automated validation check on scoring math and evidence before it is delivered.
Three-Competitor Benchmark
Your off-site signals compared side by side with three named competitors, each row ending in the action that closes the gap.
Prioritized Implementation Tickets
Work orders your developer and marketer can act on directly: the exact change, a template where one applies, an owner, the effort, and a success measure for each, in priority order.
Findings Walkthrough
A 60-minute walkthrough of the findings with JoLyn, plus a branded PDF report and an executive brief written for non-technical readers.
Re-Measurement at 90 Days
After your team implements, we re-run your exact question panel and report what moved against the baseline, and what was only normal variation.
This Is What Lands in Your Inbox
Most firms describe their audit. We would rather show you one. These are actual pages from a complete sample assessment, built for a fictional Las Vegas injury firm and rendered by the same pipeline that produces client reports, so you can judge the depth before you ever book a call.
Every factor scored on one cover: the overall score, both scoring layers, off-site authority, AI accuracy and agent readiness.
It opens on the verdict, not a compliment: how often AI names you when people don't know your name, next to Google's and Microsoft's own counts.
Who AI recommends instead of you, and the exact sources it read to decide, including who paid to appear.
Real answers from ChatGPT, Perplexity, Gemini, Copilot and Google AI Mode, each question asked three times, with cited and named kept separate.
What AI gets wrong about you, which product said it, and the listing the error came from.
A 90-day plan in three phases: every action traced to the finding it fixes and the person who owns it.
Castillo & Whitfield Injury Lawyers is fictional and the sample says so on page one; every number in it is illustrative. The sample is built in the current report format, the same one client assessments use.
How much does an AI visibility audit cost?
$500 for a small-business check on ChatGPT, $1,500 for a four-platform audit with a competitor comparison, and $5,000 for the Full GEO Audit across six AI platforms.
AI Visibility Check
Find out what ChatGPT says about your business, and the 5 fixes that matter most.
- ChatGPT tested with 10 real customer questions, each run 3 times
- What ChatGPT says about you, and whether it's accurate
- AI crawler access check (robots.txt and firewall blocks)
- Google Business Profile and key directories checked: Yelp, Bing Places, Apple Maps, BBB
- Schema basics
- 1-page scorecard with your top 5 fixes
- 15-minute recorded video walkthrough
- Delivered in 3 business days
AI Visibility Audit
Four AI platforms, your competitors side by side, and a prioritized fix list.
- ChatGPT, Perplexity, Gemini and Google AI Mode tested with 25 customer questions
- Side-by-side comparison with 3 competitors
- What AI says about you: accuracy and sentiment check
- AI crawler access check
- Google Business Profile and key directories checked
- Full schema and technical check
- Written report with a prioritized fix list
- 30-minute walkthrough call
- Delivered in 7 business days
Full GEO Audit
Six AI platforms, measured properly, with fix tickets and a 90-day re-measure.
- ChatGPT, Perplexity, Gemini, Microsoft Copilot, Google AI Mode and Grok
- 50+ buying questions, each tested multiple times with confidence ranges
- 3 competitors head to head, plus every competitor AI names instead of you, with your share of answers
- Full fact check of what AI says about you, against a verified fact ledger
- 27-factor scorecard, including off-site authority: reviews, directories, industry sources
- AI crawler access across 22 known AI crawlers
- Your own Google Search Console and Bing AI performance reports reviewed (with your access)
- Execution-ready fix tickets and an executive brief
- 60-minute walkthrough call
- 90-day re-measure included
- Delivered in about 10 business days
Upgrade credit: start with the Check or the Audit, upgrade within 30 days, and what you paid is credited in full.
Grok is tested through its developer version until consumer-app testing is available. Add Claude testing for B2B and tech companies on request (tested through its developer version).
Not sure which one fits? Run the free AI Visibility Score or book a 15-minute call.
Compare the packages
| AI Visibility Check | AI Visibility Audit | Full GEO Audit | |
|---|---|---|---|
| Price | $500, one time | $1,500, one time | $5,000, one time |
| Best for | Small local businesses | Growing businesses with real competitors | Established and multi-location businesses |
| AI platforms | ChatGPT | ChatGPT, Perplexity, Gemini, Google AI Mode | ChatGPT, Perplexity, Gemini, Microsoft Copilot, Google AI Mode, Grok |
| Customer questions tested | 10 real customer questions, each run 3 times | 25 real customer questions, each run 3 times | 50+ buying questions, each tested multiple times with confidence ranges |
| Competitor comparison | Not included | 3 competitors | 3 head to head, plus every competitor AI names, with share of answers |
| AI crawler access check | Included | Included | Included |
| Google Business Profile + directories | Included | Included | Included, plus industry directories |
| What AI says about you | Basic accuracy check | Accuracy and sentiment check | Full fact check against a verified fact ledger |
| Your Google + Bing AI reports | Not included | Not included | Reviewed |
| What you receive | 1-page scorecard + top 5 fixes | Written report + prioritized fix list | Full report, fix tickets, executive brief |
| Walkthrough | 15-minute recorded video walkthrough | 30-minute walkthrough call | 60-minute walkthrough call |
| 90-day re-measure | Not included | Not included | Included |
| Delivery | 3 business days | 7 business days | About 10 business days |
After the Audit: Implementation and Monitoring
Implementation and monthly monitoring are there if you want them, never required. Both are quoted after your audit, because they depend on what it finds.
GEO Sprint
Done-for-you implementation of your audit's fix list, 30 days
- Audit clients: a credit toward implementation within 60 days
- Your P1 tickets implemented for you
- Schema, entity graph & content fixes shipped
- Weekly progress measurement
- Wrap-up re-measure against baseline
GEO Monitoring
Your exact question panel, re-measured monthly
- Audit clients only (3-month minimum)
- Movement vs. your audit baseline
- Statistically gated, no noise-chasing
- Competitor movement tracking
- Monthly movement report
Audit delivery: about 10 business days. Every audit includes the 60-minute walkthrough and the 90-day re-measure.
What a GEO Audit Cannot Guarantee
No agency can honestly guarantee that ChatGPT, Gemini, Perplexity, Claude, or Google AI will recommend a specific company.
Those systems are built and changed by the companies that own them. Their answers vary from one run to the next, their sources shift, and none of them sells or grants a position. Anyone promising a guaranteed recommendation is promising something they do not control.
OpenAI says so directly about ChatGPT search: “ChatGPT ranks search results using multiple factors intended to help users find relevant, reliable information. Placement is not guaranteed” (Searching the web with ChatGPT, OpenAI Help Center, read September 27, 2026).
An audit also cannot:
- Tell you why a platform chose one business over another. Cited sources are evidence, not proof of cause.
- Predict what a platform will say after its next update.
- Guarantee that completing every recommendation produces a recommendation.
- Guarantee leads or revenue.
What Avante provides is documented testing, evidence, prioritized recommendations, and repeatable measurement.
You will know what AI says about your business today, with the questions, platforms, and dates on record. You will know which competitors are named and which sources are cited. You will have a prioritized plan. And at 90 days you will have the same panel measured again, so you can see what moved.
The Research and Method Behind This Service
Avante's AI Visibility Index
Our original research: 120 live queries, 60 local markets, 1,128 citations analyzed.
How AI visibility is measured
The six-level measurement framework, with definitions and a sample measurement table.
AI visibility for local businesses
What the research means for an owner, and a checklist to work through yourself.
Frequently Asked Questions
No. No agency can honestly guarantee that ChatGPT, Gemini, Perplexity, Claude, or Google AI will recommend a specific company, because nobody outside those companies controls what their systems say. What Avante provides is documented testing, evidence, prioritized recommendations, and repeatable measurement against a baseline.
A full measurement of what AI says about your business: every buying question in your market tested multiple times across ChatGPT, Perplexity, Gemini, Microsoft Copilot, Google AI Mode and Grok, with confidence ranges on every rate. An AI misrepresentation analysis checked against a verified fact ledger. A 27-factor scorecard (18 on-site, 9 off-site) weighted for your industry. A competitor benchmark. Execution-ready fix tickets with code templates, owners, and success metrics. A branded PDF report plus an executive brief written for non-technical readers. A 60-minute founder walkthrough. And a 90-day re-measure of your exact question panel, included, so you see what your fixes actually changed.
They answer different questions. The $500 AI Visibility Check tests ChatGPT with 10 questions and gives a small business its top 5 fixes. The Full GEO Audit tests six AI platforms with 50+ buying questions, each run multiple times with confidence ranges, checks every fact AI states about you against a verified fact ledger, benchmarks your competitors, delivers execution-ready fix tickets, passes an automated validation gate before delivery, and re-measures at 90 days to show what moved. If you're a single-location small business, start with the Check; anything you pay is credited if you upgrade within 30 days.
Businesses where one customer is worth real money: med spas, law firms, resorts and hospitality, luxury real estate, home services, SaaS. If AI redirecting even one customer a year to a competitor costs you more than the audit, the math works. If your average customer is worth under a few hundred dollars, start with our free AI Visibility Score instead.
About 10 business days from kickoff. Deep measurement is the reason: your question panel runs multiple trials spread across days (that's what makes the numbers trustworthy), plus the off-site authority sweep and the validation pass every report goes through before delivery.
Yes. The GEO Sprint is our done-for-you 30-day implementation, and audit clients receive a credit toward it within 60 days. None of this is required: the fix tickets are written so your own developer or marketer can execute them directly.
Yes. GEO Monitoring re-runs your exact question panel monthly and reports movement against your audit baseline, available to audit clients on a 3-month minimum. Changes are only reported as movement when they pass a statistical test, so you're never chasing noise.
Yes, and we invite it. Every finding is backed by observable public evidence, documented platform guidance, or a clearly labeled internal benchmark, with the test protocol (models, questions, dates, trial counts) documented in the report. If any finding can't be reproduced, we re-test it and correct the record.
Related Services
Related Guides & Comparisons
Before you commit, here's everything we've published that helps you evaluate the GEO Audit honestly.
GEO vs. SEO vs. Google Ads
Where a business should invest first, including when the answer is not GEO.
How to Choose a GEO Agency
Ten questions to ask any provider, and the red flags to watch for.
GEO Audit vs Traditional SEO Audit
How the methodologies differ and when you need each one.
DIY vs Hire an Agency
Honest cost comparison and when each path makes sense.
Avante vs Profound
Done-for-you audit vs enterprise SaaS platform.
Avante vs Otterly
Done-for-you audit vs a DIY SaaS tool.
What an AI Visibility Audit Costs in 2026
Honest pricing across every tier of the market.
Why You're Not in ChatGPT (10 Reasons + Fixes)
Self-diagnostic for the most common AI visibility gaps.
The 30-Point GEO Audit Checklist
Free DIY checklist you can verify yourself in an afternoon.
How to Test ChatGPT for Your Business (5 Min)
The exact manual method, no tools required.
The 30-Day GEO Implementation Roadmap
Week-by-week plan for what to ship and why the order matters.
GEO & AI Search Glossary
Plain-English definitions for every term you'll encounter.
Stop Guessing. Get Measured.
$5,000, one flat fee. Six AI platforms, documented testing, a prioritized plan, and a re-measure at 90 days. No retainer, no contract.
Questions? Email jolyn@avantevisibility.com or call (702) 350-2525.