We Checked 232 Businesses for AI Visibility. 7 in 10 Were Invisible.

Over the last few weeks, 232 business owners ran their site through our AI Visibility Check. Each one asked the same thing: when a buyer asks ChatGPT, Claude, or Perplexity for a recommendation in my space, does my name come up? The answer, for most of them, was no. 7 in 10 were completely invisible. Here is the full breakdown, what the invisible businesses had in common, and what the few visible ones did differently.

Key takeaways

• Of 232 businesses checked, 165 (about 71%) showed up zero times when AI engines were asked for a recommendation in their space. They are invisible to the people now asking AI instead of Google.
• Only 12 were genuinely winning, meaning their name came up consistently. The middle group showed up sometimes, for the wrong questions, or behind a competitor.
• The pain is not limited to one kind of business. Local service companies, coaches, agencies, ecommerce brands, and solo practitioners all landed mostly in the invisible bucket.
• The visible businesses were not bigger or older. They were clearer. Their site made it obvious who they are, who they help, and what specific questions they answer.
• AI is the new word of mouth. The businesses that get recommended are the ones the engines can actually read and place, not the ones with the most pages.
• You can check your own in about two minutes, with no guessing, at the end of this post.

How we measured this

This was not a survey. We did not ask people whether they thought they showed up in AI search. We checked.

For each business, the AI Visibility Check asks ChatGPT, Claude, and Perplexity a set of real buyer questions in that business's space. Not "is this brand good," which any engine will answer politely. We ask the questions a buyer actually types when they are looking to hire or buy: "who is the best [thing] for [situation]," "recommend a [type of provider] for [need]." The kind of question where the engine returns a short list of names.

Then we count. If the business shows up as a clear recommendation, it counts. If it shows up buried, hedged, or not at all, that gets recorded too. Eight questions per engine, across multiple engines, so one lucky mention does not inflate the result. The number you get back is how often you actually came up when it mattered.

That distinction matters. A lot of owners assume that because their site is indexed, or because ChatGPT can describe their business when you paste the URL, they are "in AI search." Those are different things. Being describable when someone already knows your name is not the same as being recommended to someone who does not.

The headline: 71% are invisible

Out of 232 businesses, 165 came back invisible. Zero recommendations across the engines for the questions their buyers are actually asking.

That is not a niche problem. That is the default state. Most businesses, including ones with real revenue, real customers, and a perfectly good website, do not exist as far as an AI engine deciding who to recommend is concerned.

The rest split like this. 47 showed up inconsistently: a mention here, a mention for a question that was not quite right there, or showing up but behind a competitor who came up first and more often. A small group, 12 businesses, were genuinely winning, coming up by name across multiple engines for the questions that lead to a sale. And 7 sat in an empty space where almost nobody in their niche showed up at all, which is the most winnable position of the four.

The shape of that result is the story. It is not a bell curve with most people in the middle. It is a cliff. Most businesses are at zero, and a small number have figured out the thing that moves them off it.

It is happening to every kind of business

One thing surprised me. I expected this to skew toward one type of business. It did not.

The 232 businesses spanned local service and trades companies (the single biggest group, things like roofers, plumbers, auto shops, clinics, med spas), coaches and consultants, marketing and web agencies, solo wellness and therapy practitioners, ecommerce and physical product brands, and professional services like legal and finance. A genuinely mixed bag.

And the invisible result cut across all of it. Being a local business did not protect you. Being an established agency did not protect you. Being a coach with a content-heavy site did not protect you. The 71% was not concentrated in one corner. It was broad.

What that tells me is that this is not about industry. A plumber and an executive coach have nothing in common except this: when their buyer asks an AI engine for a recommendation, the engine has no reason to say their name. The reason is structural, and it is the same reason for both of them.

What the invisible businesses had in common

When you look at 165 invisible sites in a row, the pattern gets loud.

They were built to be browsed by a human who already found them, not read by an engine deciding who to recommend. The homepage led with a slogan instead of a plain statement of who they help. The services were described in the language the business uses internally, not the language a buyer types. The site answered zero of the specific questions a buyer asks right before they hire. There was a Services page, an About page, maybe a blog full of announcements, and that was it.

An AI engine landing on a site like that has nothing to grab. It cannot tell, in plain terms, who this business is for. It cannot match the site to a buyer's question, because the buyer's question is nowhere on the site. So when someone asks "who should I hire for X," the engine reaches for a source that does answer that question clearly. Usually a competitor, sometimes a directory, sometimes a national brand. Anyone but the invisible business.

This is the same dynamic as word of mouth. People recommend the business they can describe in one clear sentence. "Oh you want the person who only works with X, talk to them." The business nobody can describe in a sentence does not get passed around. AI engines work the same way, at scale, instantly.

What the visible ones did differently

The 12 winners were not bigger. Several were smaller than the invisible businesses around them. They were clearer.

Three things showed up again and again. First, their site made the audience obvious. Not "we help businesses grow," but a specific kind of customer in a specific situation. The narrower it was, the more the engines could confidently match them to a question. Second, they answered real buyer questions directly, one question per page, with the answer in the first sentence instead of buried under an intro. Third, you could tell a real person or a clearly defined business was behind it, stated plainly enough that an engine could place them as the obvious recommendation for that audience.

None of that requires being a big brand. It requires being legible. The winners made it easy for an engine to say their name. The invisible businesses made it impossible, usually without realizing it.

I have run this on my own projects too. A niche site I built earned over 144,000 AI citations in a single quarter using exactly this structure, with no ads, no backlinks, and no social. The full breakdown of that build is here. The mechanism that worked there is the same one separating the 12 winners from the 165 invisible businesses in this study.

What this means for you

If you are reading this, the odds from this data are that you are in the 71%. That is not an insult. It is the default. The businesses that show up are the exception, and they got there on purpose.

The good news hiding in these numbers is that the bar is low right now. Most of your competitors are invisible too. The few who win are not winning because they out-spent anyone. They are winning because they made themselves easy to recommend before everyone else figured it out. That window does not stay open. The same thing happened with Google in the early 2000s, where being clear and early was enough, and then it got crowded. We are in the early-and-clear phase of AI recommendations right now.

The first move is not to publish more. It is to find out exactly where you stand, on which engines, for which questions. You cannot fix a gap you cannot see.

Check your own in two minutes

The 232 businesses in this study all started the same way: they ran the AI Visibility Check. You enter your site and your space, and it asks ChatGPT, Claude, and Perplexity the real buyer questions in your niche, then tells you how often you actually came up and who came up instead. It is free, and it takes about two minutes.

If you come back invisible, you are in the majority, and the fix is a known one. The AI Citation Playbook walks through the exact structure the visible businesses used, translated for an expert-led service business, for $27.

Frequently asked questions

What does it mean to be "invisible" in AI search?

It means that when an AI engine like ChatGPT, Claude, or Perplexity is asked to recommend a business in your space, your name does not come up. The engine either has nothing to say about you, or it recommends someone else. Your site can be fully indexed and still be invisible in this sense, because being findable when someone already knows your name is different from being recommended to someone who does not.

Why are so many businesses invisible to AI engines?

Because most sites are built to be browsed by a human who already found them, not read by an engine deciding who to recommend. The site does not state plainly who it helps, does not answer the specific questions buyers ask, and does not make it clear who is behind it. An engine has nothing concrete to match a buyer's question to, so it recommends a clearer source instead.

Does this only affect certain industries?

No. In this study the invisible result cut across local service businesses, coaches, agencies, ecommerce brands, and professional services alike. The cause is structural, not industry-specific. The businesses that showed up did so because their site was clear and specific, regardless of what they sold.

How do I find out if my business is invisible?

Run the AI Visibility Check. It asks the AI engines the real buyer questions in your space and reports how often your business actually comes up, on each engine, compared to competitors. It takes about two minutes and it is free.

If I am invisible, how long does it take to fix?

It is a structural fix, not a paid one, so it depends on how fast you implement. The visible businesses in this study won by making three things clear: who they help, what specific questions they answer, and who is behind the business. Those are changes you can start this week. Recommendations begin to follow once an engine has a clear, specific source to point to.