Almost every case of "ChatGPT has never heard of us" comes down to six causes: no content that answers questions directly, no schema markup, a thin entity footprint, no third-party corroboration, content rendered only in JavaScript, or robots rules that block AI crawlers outright. Each one is diagnosable in an afternoon and each has a concrete fix. Here they are in the order we find them, most common first.
Next step: see what the audit includes and costs, or request the free five-prompt snapshot.
1. Is there a page that answers the buyer's question directly?
For most invisible firms the answer is no, and this is the biggest single cause. The website describes the company: history, values, capabilities, sectors served.
Nowhere does it answer the questions a buyer actually types, so when an assistant goes looking for "can a discontinued drive range still be serviced", there is nothing on your domain for it to quote.
The fix: list the twenty most specific questions your buyers ask, and publish a page or section per question with the answer in the first two sentences, then the detail.
Question-shaped headings, plain English, real specifics. The full pattern is in how a business gets cited by ChatGPT.
2. Does the site carry any schema markup?
Usually not, which leaves machines guessing at who you are and what each page is for.
Schema is structured data embedded in the page that says, unambiguously, this is an organisation with this name, this is a service, this block is a question and its answer.
Without it, retrieval systems must infer everything from prose, and they are conservative when they are unsure.
The fix: add Organization markup site-wide, Service markup to every page describing what you sell, and FAQPage markup to the question content from point one. Validate it with a schema testing tool rather than assuming the plugin did it right.
3. Is your entity footprint more than a website?
If your company exists online as one domain and a dormant LinkedIn page, assistants have almost nothing to anchor you to.
Machines build confidence in an entity from repeated, consistent appearances: registries, directories, professional profiles, industry bodies. One thin source means low confidence, and low confidence means you get left out of answers even on topics you own.
The fix: write one canonical description of the firm, then apply it everywhere: website, LinkedIn, Companies House filings where relevant, sector directories, association profiles. Same name, same description of what you do, same locations. Boring, cheap, effective.
4. Does anyone else on the internet confirm what you claim?
Self-description is weak evidence, and assistants treat it that way. If every mention of your expertise lives on your own domain, a cautious retrieval system has no way to distinguish you from a firm that merely says the same things.
Independent corroboration, even modest amounts of it, changes how often you get named.
The fix: collect the corroboration you already qualify for: live membership listings with trade associations, manufacturer or partner pages that name you, case studies on client and supplier sites, speaker listings, the one or two trade titles your sector reads.
A handful of real, checkable mentions beats any volume of self-praise.
5. Can a crawler see your content without running JavaScript?
On many modern sites, no: the HTML that arrives is an empty shell and the content only appears after JavaScript runs.
Google renders JavaScript reliably, but several AI crawlers fetch the raw HTML and move on, so a JS-only site can rank in Google while being literally blank to some of the systems writing AI answers.
The fix: check first: view the page source and search for a sentence from your main content.
If it is missing, move the content into the served HTML through server-side rendering or pre-rendering of your key pages. Marketing pages, service pages and articles are the priority; the interactive app can stay as it is.
6. Is robots.txt blocking the AI crawlers?
More firms than you would expect are invisible because they asked to be.
In 2023 and 2024 many sites blocked GPTBot, ClaudeBot, PerplexityBot and similar crawlers by default, often via a CMS setting, a security plugin or a CDN rule nobody has revisited since. The block still stands, and the assistants respectfully see nothing.
The fix: read your robots.txt, and check your CDN's bot-management settings while you are there.
If your public marketing content is doing its job, you want the AI crawlers reading it. Allow them explicitly, point them at your sitemap, and the invisible-by-request problem is gone in a day.
What should you do with this list?
Diagnose before you spend. Run the six checks in order, note which apply, and fix the blocking issues before commissioning any content.
For specialist B2B firms the payoff stacks: unblock the crawlers, make the entity legible, then publish direct answers into a niche where almost nobody else is answering.
That sequence is the first fortnight of our AI Search Visibility Programme, and the audit that starts it exists precisely because guessing at which of the six is hurting you is slower than checking.
Quick answers
- How do I check which of the six reasons applies to my site?
- Work through them in order, they are sequenced from most to least common. Read your service pages as a stranger, check for schema with a validator, search your company name in a private browser window, look for independent mentions, view source to see if your content is in the HTML, and read your robots.txt. Most firms find two or three culprits.
- Can I fix this without rebuilding the website?
- Usually, yes. Direct-answer pages, schema markup, entity clean-up and robots.txt changes all fit into an existing site. A rebuild is only forced when the platform cannot serve content in plain HTML at all, which is rarer than agencies pitching rebuilds suggest.
- Which fix moves the needle fastest?
- Publishing direct answers to the specific questions your buyers ask, because most specialist niches have nobody else doing it. Schema and robots fixes are quicker to ship but they amplify content, they cannot replace it. Do the unblocking first, then publish.
- My competitors do none of this and still get mentioned. Why?
- Usually corroboration: they have accumulated third-party mentions, directory listings and press coverage over years, so the assistants treat them as established entities despite a weak website. You close that gap fastest on the specific questions where their content is silent, not on the broad category terms where their history wins.
See where you stand today
We ask ChatGPT, Perplexity and Google's AI answers the five questions your buyers ask, screenshot who gets named, and send you the results. Free, by hand, and no call required. If the answer is "not named", the fixed-fee audit says why.