Usually one of six things: the engine cannot reach your site, it can reach you but cannot extract a clear fact about what you do, nobody else on the internet describes you, your category words do not match what customers actually type, the model's training predates you, or it does know you and is choosing someone else. They need different fixes, so the useful first step is working out which one applies rather than doing all six at once.
"Does ChatGPT know my business" and "does ChatGPT recommend my business" are different problems with different causes, and treating them as one is why most advice on this goes nowhere.
Ask an assistant directly about your company. If it describes you accurately, it knows you, and your problem is being chosen. If it invents something, hedges, or says it has no information, it does not know you, and your problem is being read. Everything below sorts into those two piles.
The most common cause is also the most boring one. Assistants fetch pages with named crawlers, and a great many sites refuse those crawlers without anyone having decided to. Security plugins block them as unknown bots. Some hosts ship with them blocked by default.
This is worth ruling out before anything else, because no amount of writing helps a page that is never fetched, and the fix is usually a single setting. Here is how to test it in about a minute.
A model writing an answer needs a sentence it can repeat. "We help ambitious brands unlock their potential" is not one. "We do bookkeeping for construction subcontractors in Queensland, from $400 a month" is.
Most homepages are written to make a visitor feel something. Being cited requires the opposite: plain statements of who you serve, what you charge, where you operate and what you actually do, in words a stranger would use. Read your own homepage and ask what single sentence an assistant could quote. On a lot of sites there isn't one.
Assistants are cautious about naming a company on the strength of that company's own marketing. What makes a model confident is the same claim appearing on sites that are not yours: directories, comparison posts, forum threads, industry write-ups, review platforms.
This is the slowest of the six to fix because it depends on other people's websites, and it is the one that separates companies that quietly appear everywhere from companies that never do. It is also the reason a brand new site with perfect markup still gets ignored.
Companies name their category the way they discuss it internally. Customers ask in plainer language, and the assistant matches the question, not your positioning. If your site says "revenue intelligence platform" and buyers ask for "software to see which deals are going cold", the two never meet.
The test costs nothing. Write down the ten questions a customer would actually type, in their words. Search your own site for those words. If they are not there, that is the gap.
A model answering from memory alone is answering from a snapshot of the internet taken months or years ago. A company founded after that snapshot does not exist to it.
This matters less than it used to, because most assistants now run a live search before answering, which is why crawler access matters so much. But it does explain why a young company can be invisible in one engine and present in another: they are not all searching, and they are not all searching at the same moment.
The hardest version, and the one most people assume they have when they actually have one of the first five. The engine understands your company perfectly well and names two or three competitors instead, because their content answers the specific question better.
That is a separate problem with its own causes, and it is the only one of the six where the fix is writing rather than plumbing.
Run the crawler test first, because it is quick and it invalidates everything else if it fails. Then ask four engines the same customer question and write down who they name. Then read your homepage as if you were a stranger looking for one quotable sentence about what you do.
Those three take an afternoon and will tell you which of the six you are dealing with. Doing them in the wrong order is how companies spend months rewriting copy on a site no engine is allowed to read.
Open a new chat and ask the question a customer would ask, phrased the way they would phrase it: 'best accounting software for Australian tradies' rather than 'tell me about my company'. Note which companies get named. Then ask it directly what it knows about your brand. The two answers usually differ, and the gap between them is the actual problem.
No. A site can rank on page one of Google and never be named in an AI answer. Ranking rewards a page that matches a query. Being cited rewards a source that states a fact clearly enough to be repeated. Those are different jobs, and a beautifully designed site with vague copy fails the second one.
Weeks to months, and it depends on which of the causes below applies to you. Unblocking a crawler can show up in days because the engine simply could not read you before. Building the third-party mentions that make a model confident about your brand takes longer, because you are waiting on other people's websites.
Related but not the same. Most engines lean on conventional search results when deciding what to read, so ranking well makes you reachable. It does not make you quotable. Plenty of companies rank well and are still never named.
Citeon is in private beta. Tell us your domain and we'll be in touch when there's a place for you.