Executive Summary
Most small businesses are invisible to AI.
Not because they lack quality, reputation, or customers, but because AI platforms cannot find, parse, or accurately represent them. As AI-powered tools like ChatGPT, Claude, and Gemini become the primary way consumers and professionals discover businesses, a gap has opened between who your business actually is and what AI says about you.
For the majority of small businesses, AI either says nothing at all or says something wrong. Both outcomes cost you customers.
This paper introduces the Signal Strength framework and the six dimensions of AI visibility, a diagnostic lens for understanding where your business stands and what’s at stake.
At a glance
Preliminary Research Findings
What 20 small businesses revealed about AI visibility.
To test the Signal Strength framework against real-world conditions, Signal & Structure AI analyzed 20 small businesses across four industries and four U.S. cities. Businesses represent common local service categories: chiropractic practices, pilates studios, independent realtors, and bed and breakfasts. Cities included Atlanta, Raleigh, Austin, and Chicago.
Each business was evaluated against three major AI platforms: ChatGPT, Claude, and Gemini. Businesses are presented without identifying information to protect their privacy; none were clients of Signal & Structure AI at the time of evaluation.
What the data showed
Not one business in the study reached Strong Signal. The highest score recorded was 57 out of 100. The lowest was 33. The mean across all 20 businesses was 49.7, placing the group at the boundary between Low Signal and Weak Signal.
That result holds across industry and geography. The pattern is structural.
The Technical Health paradox
Most businesses scored near-perfect on Technical Health, the category that measures how accessible a website is to AI crawlers. Yet overall scores remained in the 33 to 57 range. Technical accessibility is not AI discoverability.
Structured Data: the universal weakness
Across all 20 businesses and all four industries, Structured Data was the lowest-scoring category without exception. No business scored above 48 out of 100 on this dimension, and the category average was 25. This is the markup that tells AI exactly what a business is. Without it, AI has to infer. Most of the time, it infers wrong or says nothing.
The Dataset
Twenty businesses, four industries, four cities.
Scores are on a 0–100 scale. Structured Data measures schema markup quality. Business Identity measures name, address, and phone consistency. AI Presence measures whether AI platforms found and accurately described the business.
| Business Type | City | Score | Structured Data | Business Identity | AI Presence |
|---|---|---|---|---|---|
| Chiropractic Practices | |||||
| Chiropractic practice | Atlanta | 56 | 20 | 60 | 100 |
| Chiropractic practice | Atlanta | 33 | 25 | 30 | 100 |
| Chiropractic practice | Austin | 51 | 10 | 35 | 100 |
| Chiropractic practice | Raleigh | 50 | 25 | 25 | 100 |
| Chiropractic practice | Chicago | 54 | 25 | 50 | 100 |
| Pilates Studios | |||||
| Pilates studio | Atlanta | 39 | 10 | 35 | 56 |
| Pilates studio | Raleigh | 57 | 25 | 70 | 100 |
| Pilates studio | Raleigh | 56 | 25 | 70 | 100 |
| Pilates studio | Chicago | 55 | 35 | 40 | 100 |
| Pilates studio | Austin | 49 | 45 | 30 | 83 |
| Independent Realtors | |||||
| Independent realtor | Atlanta | 48 | 48 | 5 | 70 |
| Independent realtor | Raleigh | 42 | 25 | 10 | 100 |
| Independent realtor | Austin | 52 | 38 | 35 | 100 |
| Independent realtor | Austin | 49 | 28 | 45 | 100 |
| Independent realtor | Chicago | 47 | 38 | 5 | 70 |
| Bed & Breakfasts | |||||
| Bed & breakfast | Atlanta | 53 | 38 | 70 | 83 |
| Bed & breakfast | Raleigh | 41 | 0 | 25 | 100 |
| Bed & breakfast | Austin | 40 | 0 | 30 | 100 |
| Bed & breakfast | Chicago | 43 | 10 | 35 | 80 |
| Bed & breakfast | Chicago | 55 | 35 | 100 | 67 |
By industry: pilates studios averaged 51, chiropractic practices 49, independent realtors 48, bed and breakfasts 46. The spread within each industry was as wide as the spread across industries. The difference was not the industry or the city. It was the degree to which each business had structured its information in ways AI can read.
The Problem Nobody Told You About
Your firm has 15 years of experience. AI doesn’t mention you.
A potential customer opens ChatGPT and types: “Recommend a good accountant in Austin that specializes in small business taxes.” The AI responds instantly. It names three firms. It describes their specialties, their locations, their reputations. Your firm, the one with 15 years of experience, a 4.9-star Google rating, and a waiting list of clients, is not mentioned. Not ranked low or described poorly. Simply absent.
As far as AI is concerned, your business does not exist. This is already happening millions of times a day, to businesses that have done everything right by traditional standards.
These numbers tell a clear story. AI is not replacing search; it is becoming search. If your business is not represented in AI’s answers, you are losing the highest-converting referral channel available.
The gap
The core problem is a disconnect between who your business actually is and what AI platforms say about you. For most small businesses, this gap takes one of two forms.
Invisibility. AI has no information about your business and returns nothing, or returns generic responses that omit you entirely.
Hallucination. AI presents fabricated information about your business with complete confidence. Wrong services, wrong location, invented specialties. The customer believes it. You never see the conversation.
Why Businesses Become Invisible
AI doesn’t search the way you do.
You already know how Google works: type in keywords, browse a list of links, click through, decide. AI platforms work nothing like that. They scan for structured signals across the entire web: metadata, schema markup, directory listings, review platforms, and third-party mentions. They piece together a portrait of your business from fragments scattered across dozens of sources.
This is a different game than SEO. The overlap between Google search rankings and AI citations has dropped from approximately 70% to 20% (Search Engine Land, 2025). Ranking well on Google does not mean AI knows you exist.
“Without structured data, your website is a book written in a language AI does not speak.
The five causes of AI invisibility
No structured data.
Your website speaks human, not machine. Most small business websites are built for people to read: compelling copy, clean navigation, beautiful photography. But AI platforms don’t read websites the way humans do. They look for structured data: schema markup, metadata, clearly labeled information in formats machines can parse. Without it, your website is a book written in a language AI does not speak.
Inconsistent information.
Conflicting details across the web create noise AI cannot resolve. If your Google Business Profile lists one address, your Yelp page shows another, and your website mentions a third set of services, AI faces a conflict it cannot untangle. It either picks one version (which may be wrong), tries to average them (which produces nonsense), or simply skips you.
No third-party validation.
AI trusts what others say about you more than what you say about yourself. This works the way human trust works: a friend’s recommendation carries more weight than a company’s own advertisement. AI weighs third-party mentions, reviews, directory listings, press coverage, and professional citations heavily. If the only source of information about your business is your own website, AI has limited confidence in representing you.
Thin or outdated content.
AI favors substantive, recent, well-organized information. A five-page website last updated in 2022 sends a weak signal. A business that regularly publishes detailed service descriptions, answers common questions, and maintains current information across platforms sends a much stronger one. Depth and freshness are not just good for marketing. They are how AI decides whether you are worth mentioning.
No AI-specific optimization.
The rules for being found by Google are not the same as the rules for being represented by AI. Traditional SEO focuses on keywords, backlinks, and page authority. AI visibility depends on entity clarity, content extractability, and multi-platform presence. Research shows that 40–60% of the sources cited in AI responses change month to month, but businesses with clear entity definitions and broad, consistent presence show up reliably (Search Engine Land, 2025).
The Hallucination Problem
Worse than invisibility is inaccuracy.
When AI cannot find reliable information about a business, it does not always stay silent. In many cases, it fills the gap with fabricated details presented as fact.
We tested a dental practice in a mid-size Texas city. The practice has operated for twelve years and focuses exclusively on general and cosmetic dentistry. When we asked ChatGPT what services this practice offers, it listed orthodontics and oral surgery. The practice has never offered either. When we asked Gemini, it invented a specialty in pediatric dentistry the practice does not provide.
The practice owner had no idea this was happening. There is no notification, no dashboard, no alert. The misinformation circulates in private conversations between AI and potential patients, and the business never sees it.
From the Founder
When we launched Signal & Structure AI and tested our own AI presence using our proprietary Signal Score methodology, we scored 0 out of 100.
Over the next 60 days, as we applied the same framework we use with clients, our score climbed: 0, then 6, then 76, then 75, and most recently 80 out of 100, reaching Strong Signal. AI platforms now mention us accurately.
We were a new business, so a zero on day one was not surprising. What is surprising is that businesses with ten, fifteen, twenty years of operation, hundreds of reviews, and strong local reputations score the same way. Our zero was a starting point. Theirs is an invisible crisis they do not know they have.
The Signal Strength Framework
A thermometer for AI visibility.
To fix a problem, you first need a way to measure it. Most business owners are trying to solve AI visibility without a thermometer. Signal Strength is a measure of how accurately and completely AI can represent who you are, what you do, and why you matter.
| No Signal | AI cannot find or describe the business. Responses are blank, generic, or entirely fabricated. | "I don’t understand why nobody finds us online." |
| Low Signal | AI mentions the business but gets major details wrong. Wrong services, wrong location, fabricated information presented as fact. | "People keep asking about services we don’t even offer." |
| Weak Signal | AI has basic information correct but significant gaps exist. Some platforms know you, others don’t. Differentiators are missing. | "The basics are there but it doesn’t capture what makes us different." |
| Strong Signal | AI accurately represents the business across platforms. Identity, services, positioning, and differentiators are correct and consistent. | "When someone asks AI about us, they get the truth." |
In our preliminary testing, most small businesses fall in the No Signal or Low Signal categories. Strong Signal is not theoretical. It is the difference between a potential customer asking AI for a recommendation and hearing your name, or hearing your competitor’s.
Six Dimensions
What we measure when we measure your signal.
Signal Strength is not a single measurement. It is the combined result of how well your business performs across six distinct dimensions, tested across ChatGPT, Claude, and Gemini.
If AI cannot correctly name your business, describe what you do, or distinguish you from competitors, every other dimension fails. Identity is the foundation.
Misrepresented services mean mismatched referrals. When AI tells a customer you offer something you don’t, or omits the service they need, the opportunity is lost before it starts.
A correct but generic description makes you interchangeable with every competitor. AI needs to understand your differentiators, not just your category.
For service-area and brick-and-mortar businesses, location accuracy directly determines whether AI refers local customers to you or to a competitor across town.
Reviews, awards, professional recognitions, and third-party endorsements carry disproportionate weight in AI’s decision-making. Without them, you lack the social proof AI uses to rank recommendations.
The ultimate test. When a potential customer asks AI for a recommendation in your category and location, are you in the answer? This dimension measures whether all other signals combine into actual visibility.
A physical therapy clinic with over 200 five-star reviews scored strong on reputation but had zero discoverability. A landscaping company with detailed service pages scored well on services and identity, but AI consistently placed them in the wrong zip code because of conflicting location data across directories. The framework is designed to pinpoint exactly where your AI visibility breaks down, not just whether it exists.
What This Means for Your Business
The cost of invisibility.
AI invisibility is not an abstract technical problem. It creates three specific, measurable impacts on your business.
The compounding referral gap.
Every week that your business is absent from AI recommendations, potential customers are being routed to competitors. If AI handles even five recommendation conversations per week in your category, that is 260 missed opportunities per year. Unlike a bad Google ranking, which you can see and track, these losses are completely invisible to you.
The silent customer loss.
When AI fabricates details about your business, the damage happens quietly. A customer reads that you offer a service you do not provide, visits your website, sees no mention of it, and moves on. They do not call to ask. They simply disappear.
Competitive displacement that accelerates.
AI platforms learn and reinforce over time. A competitor that appears in AI answers today will appear more frequently tomorrow. The gap between visible and invisible businesses widens over time. Waiting is not neutral; it means falling behind.
The first-mover advantage.
Here is the good news: this is early. Most businesses have never asked “What does AI say about me?” The concept of AI visibility is new enough that the playing field has not yet been defined.
Research from Princeton, Georgia Tech, and IIT Delhi found that adding citations, statistics, and expert context can improve AI search visibility by 30–40% (2024). The window for first-mover advantage will not stay open indefinitely. The time to act is before your competitors realize they need to.
Finding Out Where You Stand
Two ways to start.
Signal Pulse is a free quick check available at signalstructure.ai. It evaluates two of the six dimensions — enough to tell you whether you are broadcasting or invisible. It takes minutes and costs nothing.
For the complete picture, the founding beta is where Signal & Structure AI evaluates your business across all three major AI platforms and all six dimensions. You receive your score, the story AI is telling about you, what’s accurate, what’s wrong, what’s missing, and a clear path to improve it.
You’re not broken. You’re just not broadcasting. That’s fixable.
References
Sources cited in this paper.
- Gartner. (2026). “Gartner Predicts Search Engine Volume Will Drop 25% by 2026, Due to AI Chatbots and Virtual Agents.” Gartner Press Release.
- Google. (2026). “AI Overviews: Expanding Access and Improving the Search Experience.” Google Blog, The Keyword.
- OpenAI. (2026). “ChatGPT Reaches 800 Million Weekly Active Users.” OpenAI Blog.
- Previsible. (2025–2026). “AI Referral Traffic Conversion Rates: How LLM-Driven Visits Compare to Organic Search.”
- Princeton University, Georgia Tech, & IIT Delhi. (2024). “GEO: Generative Engine Optimization.” Research Paper.
- Search Engine Land. (2025). “The Overlap Between Google Rankings and AI Citations Has Dropped from 70% to 20%.”
- Semrush. (2025). “The State of LLM Referral Traffic: 800% Year-Over-Year Growth in AI-Driven Website Visits.”
Signal Score is a proprietary methodology of Signal & Structure AI. This is a preliminary edition; methodology, dataset, and findings will be updated as additional businesses are added to the study.