The definition, in one paragraph
AI visibility is how often, and how favorably, your business shows up when people ask an AI assistant a question you would want to be the answer to. Showing up means one of three things. The assistant names you. It links to your website as a source. Or it recommends you, ahead of the others it mentions.
That is the whole idea. A person types or says something like "who is a good electrician near me" or "what is the best way to store cast iron" into ChatGPT, Gemini, Claude, Perplexity or Google's AI answers. The assistant writes a short reply. Some businesses are in that reply. Most are not. The ones that are in it have AI visibility for that question. The ones that are not, do not.
Notice what is missing from the definition: a ranking, a score, a tool. Those come later, when people try to measure it. At bottom, visibility is a plain fact about a particular answer on a particular day. You are in it, or you are not.
If you want to see the fact for your own business, the page on checking your own business explains how to look.
A simple example
Imagine you run a small store that sells aquarium plants. A customer wonders which plants are easy for a beginner, opens an assistant and asks. The assistant writes a short answer listing five plants and mentions two stores where they can be bought, with a link to each.
If your store is one of those two, you have visibility for that question. A reader may click, or may simply remember your name. If your store is not, then for that person at that moment you did not exist, however good your plants are.
Now multiply that by every question your customers might ask. Which plants suit low light. Where to buy plants that arrive alive. How to set up a small tank. Each is a separate chance to be in the answer, and your visibility is the share of those chances you actually capture.
That share can be low even for a good business with a well-ranked website. Visibility in an assistant is not the same as ranking in Google, which is the next thing to understand.
How it differs from Google rankings
For two decades, being found online meant ranking: climbing a list of links so that more people clicked. You held a position, and you could see it.
Assistants work differently. They read many sources and write one answer. There is no list of ten positions, and nothing to climb. The reader often never sees the sources unless they click. Your goal changes from holding a position to being included, described well, and chosen.
This does not mean search has stopped mattering. Google is still where a large number of buyers begin, and good pages that rank well are often the same pages that assistants draw on. Search is changing, and a business needs to be findable in both places. The two overlap, but they are not identical. A page can rank well and never be cited. A business can be named in assistant answers because of a forum thread or a review site, even if its own website ranks poorly.
Our comparison of SEO, GEO and AEO sets out how the disciplines relate. In short: an LLM, or large language model, is the kind of AI behind ChatGPT, Claude and Gemini. GEO means generative engine optimization, which is work aimed at being included in the answers such models write. AEO means answer engine optimization, which is the same aim stated from the viewpoint of any system that gives direct answers. All three names circle the idea of AI visibility.
Why it matters now
Three plain reasons, none of which need a forecast.
People already ask. Many buyers now put a question to an assistant before, or instead of, typing a search. They ask for a shortlist, a comparison, or a recommendation. If your business never appears in those answers, you miss the people who ask that way.
The answer is short. A search page has room for ten results. An assistant often names only a few businesses. Being third in a list of ten is fine. Being sixth in a list of three is impossible, because there is no sixth place. The spots are fewer, so being left out costs more.
The reader trusts the answer. People tend to treat an assistant's reply as an informed opinion rather than a set of links to evaluate. When it names a business, that carries weight. When it does not, the reader has no reason to look further.
None of this calls for panic. It calls for knowing where you stand, which is cheaper to find out than most people expect.
What decides who gets named
Nobody outside the companies knows the full recipe, and anyone who claims to is guessing. But the broad pattern is observable, and we cover it in detail in our write-up of how AI assistants pick businesses, which includes our own field test. The short version is that assistants tend to name businesses that many independent sources describe, clearly and consistently, for the specific job asked about.
That tilts the field in a way many owners find surprising. A large brand with a big advertising budget does not automatically win. A smaller business that is well documented by customers, local sources and its own clear pages can be the one named. Clear, specific, honest information that other people repeat is the common thread.
It follows that visibility is not a trick or a setting. It is mostly the by-product of being easy to describe accurately, in many places, in the words buyers use.
What AI visibility is not
Some clarifications, because the term gets stretched.
It is not traffic. A recommendation can send no visitors at all and still shape a decision, because the buyer remembers the name and searches for it later. Traffic is one signal among several.
It is not a single score. Tools sometimes report one number. Behind it sits a set of answers, and the number can hide a lot. Our page on the way to measure it explains what to count instead.
It is not guaranteed or permanent. The same question can give different names on different runs, and assistants change over time. Visibility is a pattern you maintain, not a position you win once.
It is not only for big companies. If anything, local and specialist businesses often do well, because the questions people ask are specific and the sources about them are close at hand.
It is not a replacement for good work. Assistants describe what the world says about you. If the underlying business, reviews and information are weak, no amount of effort fixes that in the answer.
Where a beginner should start
Do these three things, in this order, and you will understand your position better than most businesses do.
First, ask. Open two or three assistants in a logged-out window and ask the five questions your best customers would ask. Read the answers. Write down who is named and whether you are in them.
Second, read how you are described. If you are named, notice the words used. They are a live summary of what the assistant has picked up about you. If they are wrong or dated, you have found something fixable.
Third, find out why a competitor appears. Open the sources the assistant links to. What do those pages have that yours do not? A clear answer near the top, a comparison table, customer quotes, a price? That is your first piece of homework.
If you are missing from nearly everything, our page on why a brand is not in ChatGPT covers the usual causes. And a word of caution: run each question more than once, because one answer proves little.
Three levels of visibility
It helps to think of visibility as a ladder with three rungs, because the rungs are worth different amounts.
The bottom rung is being known. The assistant has enough information to say what your business is, where it operates and what it does. You can test this by asking about your business by name and reading what comes back. If the answer is vague or wrong, higher rungs are out of reach for now.
The middle rung is being included. When a buyer asks a general question with no name in it, you appear in the answer. This is the rung most people mean when they say visibility, and the hardest to reach, because you are competing with every other business the assistant could have named.
The top rung is being chosen. You are named first, or the assistant says outright that you fit the situation described. This is where visibility starts to look like sales.
Most businesses begin on the bottom rung for some questions and are absent from the rest. That is normal. Progress is not one jump. It is a slow filling in, question by question.
Who controls it, and who does not
A common worry is that visibility is out of your hands. Partly it is, and partly it is not, and it is useful to see the split.
You control your own pages. Whether they answer questions plainly, say who you serve, give real details and stay current. You control whether the facts about you are the same everywhere you publish them. You control how you treat customers, which feeds the reviews that others write.
You do not control what a review site, a forum or a publication says. You do not control which sources an assistant prefers, or when it changes its mind. You cannot pay for a place in an answer the way you can pay for an ad slot beside it.
The sensible stance is to do the part you control very well and to earn the part you do not, by being the kind of business others want to describe accurately. It is slower than buying a slot, and it lasts longer.
What buyers actually ask
It is easy to imagine visibility as being found by your name or your category. In practice the questions are more specific and more varied than most owners expect.
Buyers ask for a shortlist for a particular situation. They ask which of two options suits a given budget or problem. They ask what to look for before hiring or buying. They ask whether a named business is any good. They ask how a thing works before they know they need your service at all.
Each type of question favors a different kind of page. A shortlist question favors sources that compare several providers. A should-I-hire question favors plain explanatory pages. A is-this-business-good question favors reviews and mentions in independent places. So visibility is not one target. It is a spread of small targets, and a business usually does well on some and poorly on others.
That is why a single check tells you so little and a set of questions tells you a great deal. The set shows where the gaps are, and the gaps tell you what to write or earn next.
Words you will meet
A short glossary, so the vocabulary does not get in the way.
Assistant. A chat-style AI that answers questions, such as ChatGPT, Gemini, Claude or Perplexity.
Mention. Your name appears in the answer. It says nothing yet about whether the mention is good.
Citation. The answer links to a source, often your own page. We keep a plain entry on what an AI citation is.
Recommendation. The assistant tells the reader to pick you, or lists you first as the best fit.
Prompt. The question or instruction a person types. In this field the prompts you test are your buyers' questions.
Share of answers. The portion of tested answers in which you appear. It is the basic unit of measurement.
You can read all the pages on this site without any other vocabulary. Where a technical term is unavoidable, it is explained where it first appears.
How RunOctopus approaches it
RunOctopus began with a pet supply store, All Angles Creatures, which went from no Google clicks to more than 4,500 a month in five months. The company now does the same kind of work for other businesses, as a done-for-you service.
Ollie, its AI, researches a business and then writes and publishes pages on the business's own website, with a team that plans, writes, reviews and publishes alongside it. It works with Shopify, WordPress, WooCommerce and Wix, and there is nothing to install or approve. Customers also get reports on where ChatGPT, Perplexity, Claude and Gemini name them. The company does not offer a free AI visibility checker.
If you are curious what Ollie would plan for your site, the free site read at app.runoctopus.com/start reads your address and shows the plan. It does not test AI answers. No account, no call.