The AI Visibility Index: Measurable ChatGPT's Recommendations
December 28, 2025
LLMs' answers are non-deterministic. If you ask ChatGPT the same question twice, you'll get two different responses.
For businesses, this creates a serious problem as it introduces a lack of predictability.
A law firm might be recommended today, ignored tomorrow, and replaced by a competitor the day after - with no visible explanation. Marketing and sales teams are left guessing whether they are winning, losing, or just experiencing randomness.
The AI Visibility Index exists to solve this exact problem.
The Core Problem: AI Is Non-Deterministic by Design
AI systems like ChatGPT do not operate like Google search rankings. There is no single "position #1."
Instead, results vary by:
- User's profile and persona
- Query phrasing and context
- Time of day and model version
- Sources' contents and availability at the moment of search
A single answer tells you almost nothing.
Yet today, most businesses still evaluate AI visibility based on one prompt, one screenshot, one moment in time. That approach is fundamentally flawed.
Why a Single ChatGPT Answer Is Meaningless
Imagine judging stock performance based on one trade. Or evaluating SEO based on one Google search.
That's effectively what most businesses are doing with AI today.
The real question is not: "Did ChatGPT recommend us?"
The real question is: "How often does ChatGPT recommend us over time?"
This distinction changes everything.
Introducing the AI Visibility Index
The AI Visibility Index is a statistical benchmark that measures how AI systems recommend businesses across hundreds of independent AI search sessions.
Instead of observing one answer, we observe patterns.
Each Index is built by running 250+ independent AI searches across different times, phrasings, and contexts for the same industry and location (e.g., Divorce Lawyers in New York).
From this data, we calculate recommendation frequency, ranking stability, volatility, and source dependency.
The result is not an opinion. It's a probability distribution.
What the AI Visibility Index Reveals
The AI Visibility Index turns AI signals into something you can predict and optimize. It answers questions like:
- Are we consistently visible, or just occasionally lucky?
- Is our position improving or eroding?
- Where should we invest to increase or protect our AI visibility?
The 9 Metrics Inside Every AI Visibility Index
Each AI Visibility Index contains nine distinct metrics that together paint a complete picture of your AI presence. Here is what each one measures and why it matters.
1. AI Recommendation Rate
This is the most fundamental metric: out of all the AI searches we ran, what percentage included your business in the recommendations?
A business appearing in 65% of searches has a vastly different market position than one appearing in 12%. This metric separates consistent performers from occasional mentions.
2. AI Market Share Score
Not all mentions are equal. Being recommended first carries more weight than being listed fifth.
The AI Market Share Score is a weighted calculation where higher ranking positions earn more points. A business consistently recommended in position #1 will significantly outscore a competitor who appears more often but always in lower positions.
3. Source of Truth Stability
This metric reveals which domains ChatGPT cites most frequently when making recommendations in your industry. It answers: "Where is the AI getting its information?"
If ChatGPT consistently cites Yelp, ZocDoc, or a specific industry directory, you know exactly where to focus your presence optimization efforts.
4. Citation Impact Score
While Source Stability shows frequency, Citation Impact measures influence. Some sources are cited repeatedly across many searches; others appear once and never again.
This weighted score helps you understand which citations carry the most weight in shaping AI recommendations for your market.
5. Aggregator Impact Score
Aggregator platforms like Yelp, TripAdvisor, ZocDoc, and Healthgrades play a special role in AI recommendations. They serve as third-party validation that the AI uses to verify business information.
This metric isolates aggregator influence specifically, showing which platforms drive the most AI visibility in your industry and location.
6. Primary vs. Secondary Sources
Primary sources are official business websites. Secondary sources are aggregators, directories, and third-party mentions.
This ratio reveals whether the AI is trusting your own content or relying entirely on what others say about you. A heavy skew toward primary sources (75%+) might seem good, but it actually indicates risk - it means you're not present on the aggregators that AI uses for validation.
7. Source Category Breakdown
Not all sources are aggregators or business websites. AI citations span multiple categories: news articles, review sites, industry registries, government databases, social media, and more.
This breakdown shows the full distribution, helping you understand the entire ecosystem of sources that influence AI recommendations in your market.
8. AI Query Variants
Users don't all ask the same question. Some search for "best dentist in Manhattan," others ask "top-rated dental clinic near Central Park," and others simply want "dentist recommendations NYC."
This metric shows the actual query variations ChatGPT used during its search process, revealing how users phrase their requests and which phrasings your business appears in most often.
9. Website Verification Rate
When ChatGPT recommends a business, does it visit the business website to verify information, or does it recommend based solely on aggregator data?
This metric tracks the AI's verification behavior. A high "direct" rate means recommendations are made without visiting your site - which means your website content isn't influencing the decision. A high "verified" rate means the AI is actively checking your site before recommending.
From Randomness to Strategy
ChatGPT is non-deterministic. That fact will not change.
But non-deterministic does not mean unpredictable. With enough observations, patterns emerge. Market leaders become visible. Source dependencies become clear. Optimization opportunities become obvious.
The AI Visibility Index turns randomness into measurable, predictable insights you can use to show up in ChatGPT answers - not by luck, but by strategy.
Frequently Asked Questions
How many searches does each AI Visibility Index include?
Each Index aggregates 250+ independent AI searches conducted across different times, query phrasings, and contexts. This sample size is large enough to reveal statistically meaningful patterns while filtering out random noise.
How often are AI Visibility Indexes updated?
We publish new Indexes periodically as we collect fresh search data. Each Index represents a specific time period (typically one month) to show how AI recommendations evolve over time.
Can I get an Index for my specific industry and location?
We are continuously expanding our coverage. If you need an Index for a specific industry-location combination that is not yet available, contact us or run an individual AEO Report for your business.
What is the difference between an AI Visibility Index and an AEO Report?
An AI Visibility Index shows market-wide data for an entire industry in a location - who dominates, which sources matter, and overall patterns. An AEO Report focuses on your specific business, showing exactly how ChatGPT reasons about you and what you can do to improve.
Explore Available Indexes
We publish AI Visibility Indexes for specific industries and locations. Each Index aggregates hundreds of real AI searches to reveal which businesses dominate ChatGPT recommendations - and why.
Browse available indexes to see how your industry and location performs, or run your own AEO Report to see exactly where your business stands in the AI recommendation landscape.
Check your business visibility in AI
Generate an AEO report and discover what ChatGPT says about your industry.
Generate Report