Prospective students are asking an assistant about you. Can it read your site?
Most universities look healthy in Google and are invisible in ChatGPT, Gemini and Perplexity. The usual cause is not content. It is a bot rule at the edge, a robots.txt line, or course detail that only exists after JavaScript runs. This is the only audit we run: all seven criteria, from our servers, with real crawler User-Agents. Give your work email and the full report becomes a PDF download on this page.
One audit, seven criteria
- Your headline read on this page: green, amber, red or grey.
- The single highest-cost issue, in plain language.
- All seven criteria with the evidence behind each one, downloadable here.
- Your result next to the anonymized sample: the share of readable peers that pass each criterion, by sector, by region if you tell us, and by platform where we fingerprint one.
- Findings ranked by what to change first.
Cohort sizes are stated next to every comparison. Where fewer than five peers could be read, we say the cohort is too small instead of quoting a number. No institution is ever named.
What the audit looks at, and why each one decides whether an assistant can quote you
01
AI crawler policy
Which AI crawlers your robots.txt allows, split by answer engine and training bot.
02
Edge and WAF behavior
We replay your homepage as OAI-SearchBot, PerplexityBot and GPTBot to see if the edge blocks or challenges them.
03
Structured data
Whether CollegeOrUniversity, Course, BreadcrumbList and FAQPage schema exist and parse.
04
Course discovery
Whether course pages are in a sitemap and return readable server-rendered HTML.
05
PDF dependency
Whether fees, entry requirements and prospectus content are locked in PDFs.
06
llms.txt
Whether you publish an llms.txt map of the pages you want quoted.
07
Platform fingerprint
The CMS and edge stack, so fixes land in the right place.
Why is a university invisible in ChatGPT but fine in Google?
Google renders JavaScript and has crawled the site for years. Answer engines fetch far fewer pages, rarely execute scripts, and are often stopped earlier: a robots.txt line that names GPTBot, a bot-management rule at the edge that challenges OAI-SearchBot, or course detail that only appears after a client-side fetch.
Does blocking AI crawlers protect content?
It separates two different bots. Training crawlers collect data for model training. Retrieval crawlers such as OAI-SearchBot and PerplexityBot fetch a page at the moment a prospective student asks a question, and that fetch is how you get cited. Blocking the retrieval bot removes you from the answer without changing training exposure.
What does the audit actually request?
Public pages only: robots.txt, sitemaps, the homepage, a sample of course pages and llms.txt, requested from our servers at a polite rate with real crawler User-Agents. It never submits forms, never signs in, and writes nothing to your site.
What do you get at the end?
A downloadable report with all seven criteria and the evidence behind each one, your result next to the anonymized sector sample, and the findings ranked by what to change first, so your platform team can reproduce every result from your own network.