Language schools and immersion programmes
Three in ten language schools carry a critical fault.
Language school decisions are made by students and parents researching from another country, in another language, increasingly by asking an AI assistant. That is the one audience that cannot phone you to check.
Languages measured: Spanish, Mandarin, Portuguese.
We ask in the language your students actually search in, not in translated English — and we include a model used in mainland China, where ChatGPT is not available. Most tools measure neither.
124 language schools, audited.
- 124
- audited
- 30%
- carry a critical fault
- 19%
- have a server refusing an AI crawler
- 20%
- publish an llms.txt
Median readiness score 71 out of 100. 7% block a crawler deliberately in robots.txt — more than any other programme type. Audited 8 August 2026.
The highest critical-fault rate of any large category.
Of 124 language schools we audited, 30% carry a critical technical fault, against an average of 27% across educational travel — and the highest rate of any category with more than a hundred organisations in it.
They also block AI crawlers deliberately more often than anyone else: 7%, against 1–5% everywhere else. Some of that is intentional and considered. Much of it is a plugin default nobody chose.
Free score. A person runs it, not a robot.
Enter your website address and we will run the technical audit on your real site — every AI crawler tested, twice. You get your score and the single biggest thing holding you back, within one week.
No account, no card.
Four out of five sites we audit have at least one fault we can name. The typical site has two or three. The free score gives you your score and the most serious one, in plain English, with the fix. The rest come with the full audit — along with the half of the picture the free score cannot see: whether AI assistants actually recommend you, and who they recommend instead.
What the full audit covers
- 60 real questions from students and parents, across 5 AI models in Spanish, Chinese and Portuguese.
- Every question asked three times.
- Whether AI recommends a school at all, or points students to Duolingo and italki instead.
- How it answers the accreditation and student-visa questions that decide enrolments.
- Which schools get named in your place, in each source market.
- Which directories and review sites the models actually trust.
- Eight prioritised fixes.
- Written recommendations from me, not a dashboard export.
The five assistants we ask: ChatGPT, Claude, Perplexity, Gemini and Qwen. The first four search the web while they answer; Qwen answers from what it already knows, and is included here rather than a European model because a large share of language-school enquiries begin in Chinese.
Read what we can and cannot measure, and why we report the limits — our method →
Find out what AI tells a student in Bogotá, São Paulo or Shanghai.
Start with the free check above, or ask us a question.
The audit is $950, or $712 with an annual plan. See what’s included →