Decided elsewhere
- Reviews and directories
- News, forums and comparison articles
- How often you are named in answers
Half of it you control, and the audit reads that half in under a minute. The other half is explained below, without the sales pitch.
The part of the answer that lives on your own website, which is the part you can do something about today.
ChatGPT sends three differently named agents. We check your rules against all three.
We have an AI read the page the way an engine would, then report what it could not name.
Text that waits for JavaScript, or a headline that is a slogan, leaves nothing to lift.
When ChatGPT names a business it is mostly repeating what other sources say about it.
GPTBot
Collects pages for training. Blocking it keeps you out of the model itself.
OAI-SearchBot
Builds the search index behind ChatGPT search. This is the one that matters most today.
ChatGPT-User
Fetches a page live when someone asks ChatGPT to look at a link.
Blocking all three is a decision some publishers make deliberately. Blocking them by accident, through an inherited robots.txt rule or a firewall setting nobody remembers, is far more common.
Weighted the way an engine that reads pages rather than crawling the live web would, with ChatGPT’s three crawlers checked separately.
GPTBot, OAI-SearchBot and ChatGPT-User, apart
Schema, heading order, sitemap, llms.txt
What a model can name from the page alone
Mobile performance, and how fast the page answers
All of it on screen the moment the scan finishes. No summary version, no upgrade.
AI readiness score
An AI engine gets about half of what it needs from this page.
Also in it:
0 to 100, from 4 weighted signals, with a grade and a band.
Five of them: crawler access, structure, understanding, speed, security.
12 agents across 6 engines, checked one at a time.
Five facts: your business, services, products, industry and location.
Every one, with what it is worth and how much work it is.
Four: ChatGPT, Claude, Gemini and Perplexity, weighted separately.
One page, read live, with your ChatGPT readiness score at the top and the full audit under it.
Start at crawler access.
ChatGPT sends three differently named agents and we check each one. Blocking them by accident is far more common than blocking them on purpose.
Start at what a model could name.
This is the half you control. A page that never says what the business does gives a model nothing to repeat.
Send the report link.
It separates what you control from what you do not, which is usually the argument being had.
It recommends businesses it can describe confidently from sources it trusts. That starts with your own site being readable and clear about what you sell, and it ends with what other sites say about you. This check covers the first half, which is the half you control.
Because it mostly repeats what other websites say about each of you, not what your own site says. Being readable is the floor, not the ceiling. The guide on appearing in ChatGPT covers what actually moves it.
It draws on what it was trained on and, in search mode, on pages it fetches at the time of the question. Both favour businesses that are described consistently in plain language across the web, with a site that confirms rather than contradicts the description.
Both, and the second one carries more weight than owners expect. Your site decides whether you are eligible and what the facts are. Directories, reviews, press and anywhere else you are described decide whether you come up at all.
No, though they correlate. Google returns a list and you hold a position in it. ChatGPT returns one answer naming one or two businesses, so there is no second place, and a page that ranks well but reads vaguely can be skipped entirely.
That is a real decision, not an obvious one. Blocking it keeps your content out of training but does not keep you out of ChatGPT search, which uses OAI-SearchBot. Publishers often block the first and allow the second. Every crawler is listed here.
Unblocking a crawler takes effect at the next crawl, usually days, and ChatGPT search picks changes up fastest. Changing what the model says about you is slower and nobody can promise a date, because it depends on sources you do not control.