If you ask ChatGPT or Perplexity to recommend a business like yours and your name never comes up, the reason is usually one of three things, and none of them is that your website is not good enough. Either the crawlers cannot reach your site, or the model cannot tell what your business actually is, or there is nothing on your pages written plainly enough to be quoted. All three are fixable, and the first one is often fixed in a single setting.
First, check that the door is open
Open yourdomain.com/robots.txt in a browser and read it. It is a plain text file and it is public on every website.
Look for crawler names with Disallow underneath them. If the AI crawlers are listed there, your content is not being read, and nothing that is not read is ever cited.
This happens by accident far more often than on purpose. Hosting platforms and content delivery networks have added settings that block AI crawlers by default or behind a switch nobody explained, and the site owner finds out months later.
Security rules cause the same thing more quietly. A firewall or rate limit that returns an error to anything that is not a normal browser will refuse the crawler without ever appearing in robots.txt.
There is a particular irony worth avoiding here: a business that sells AI visibility while blocking the AI crawlers on its own site. Anyone who checks can see it in ten seconds.
Second, make it obvious what your business is
A model recommends a business when it is confident about what that business is, where it is, and what it does. Confidence comes from seeing the same facts stated the same way in several places.
Structured data is the most direct way to state them. An organisation block with a name, address, phone number, opening hours, coordinates and a link to your Google profile turns your details into facts rather than sentences that have to be interpreted.
Links to your own profiles elsewhere matter more than they look. Listing your LinkedIn, your Google Business Profile and your other accounts in your structured data is how you confirm that all of them are the same organisation. Leave that empty and every mention of you is a separate guess.
Then check that those places agree with each other. A different address on a directory, an old phone number on a profile, a business name with an extra word in it: each disagreement lowers confidence a little.
Third, write things that can be quoted
Models quote passages, not pages. A passage that can only be understood by reading the two paragraphs above it cannot be lifted, so it never is.
The practical version of this: put the answer in the first two sentences under each heading. Then explain. The explanation is for the human reader, and the opening is what gets quoted.
Write the questions your customers actually ask, in the words they use, and answer them directly. Not the polished version of the question, the real one.
Specific beats general every time. A page that names actual figures, actual steps and actual local detail gives a model something to say. A page of confident generalities gives it nothing it does not already have from a thousand other sites.
How to test it properly
Ask the tools the questions your customers would ask, not questions about your company. Nobody types your business name into ChatGPT before they have heard of you.
Ask a service and a place: a dentist in a named area, a plumber for a named problem. See which businesses get named, and look at what those businesses have that you do not.
Do it in several tools, because they do not agree. Being named by Perplexity and ignored by ChatGPT is common, and it usually points at a crawler problem rather than a content one.
Write down what you find before you change anything. Without a baseline there is no way to tell later whether the work did anything.
Questions people ask
- Does blocking AI crawlers protect my content?
- It stops that crawler reading the page, and it also stops it citing you. There is a real argument for blocking training crawlers if you do not want your writing used that way. Just know that the same setting often removes you from the answers your customers see, so it is a trade rather than pure protection.
- How long until AI starts mentioning us?
- Once the crawlers can read the site and the structured data is in place, citations often start moving in four to eight weeks. It is usually faster than ranking movement, because there is no queue of competitors to climb past, only a question of whether you are legible.
We do this
If you would rather have it done than do it, this is the part of what we offer that covers it.