If you want to know whether AEO for language schools is working, ask ChatGPT which language school in your city is best and see whether your name appears. That one query answers the question faster than any audit.
For most schools the name does not appear. What appears instead is a directory listing, a large chain, a university extension program, and a careful paragraph explaining that the answer depends on the student’s goals. Your school, which may well be the best option in the city, is absent from the conversation where the decision is now being made.
This is a different problem from search ranking, and it does not get solved by the same work. Traditional search rewarded pages that matched a query. Answer engines reward pages that resolve a question, from a source they can attribute, corroborated somewhere other than your own domain. Language schools are unusually badly positioned for this, because the category’s website convention is to be encouraging and non-committal about exactly the things a person needs to know.
Assistants cannot use a page that refuses to commit
Open your own levels page. It probably says that progress depends on the individual, that our experienced teachers will help you reach your goals, and that classes are small and supportive.
Every sentence there is true and none of it is usable. An assistant asked how long it takes to reach conversational Spanish cannot extract an answer from a page that declines to give one. So it goes to a source that does, which may be a competitor, a blog, or a general article that has nothing to do with your city.
The category writes this way for understandable reasons. Committing to a timeline feels like a promise you cannot keep, because students vary enormously. Naming a price invites comparison. Saying who your program suits badly feels like turning away business. So the page hedges, and the hedge makes the page invisible to the systems now doing the recommending.
The way out is to commit with conditions rather than refusing to commit. Not that it takes four months. That it takes roughly four months at four hours a week for a learner with no prior Romance language exposure, and roughly ten weeks for one who studied French at school, based on what you have observed across your own intake. That sentence is specific, defensible, honest about variance, and extractable. It is also something almost no competitor in your city has written down.
It helps to think about what the assistant is doing when it answers. It is not ranking your page against a competitor’s. It is trying to construct a paragraph that will satisfy a person, and it needs sentences it can stand behind. A page full of reassurance gives it nothing to build with. A page with four concrete figures and a stated method gives it the whole paragraph, and the attribution comes with it.
This also explains why some schools with weak traditional rankings get named often. Search ranking rewards domain strength, link volume, and years of accumulated authority. Answer construction rewards whoever wrote the clearest usable sentence. A four-page site that commits to specifics can beat a large chain’s vague landing page in an assistant’s answer while losing to it badly on a search results page.
There is a secondary effect worth noticing. The discipline of writing committed answers improves the school. Once you have published observed timelines you start measuring them, and once you measure them you find the pacing problems. Several of the operational improvements schools credit to their teaching came from being forced to state a number publicly.
One more thing about hedging. The instinct to protect yourself legally by refusing to state timelines is usually misplaced. A published range with stated conditions and an explicit note that individual results vary is more defensible than a vague promise of success, because it describes a method rather than guaranteeing an outcome. Ambiguity feels safe and is not.
The six answers, and how to write each one
There are six questions that carry almost all the decision weight in this category. Each deserves its own page, written to be quoted.
The first is how long each level takes, in hours of instruction and in calendar weeks, at each of the schedules you offer. Give the range and the conditions that move it. State what you count as a level, because the frameworks differ and a reader comparing two schools has no idea whether your intermediate matches theirs. If you teach to the Common European Framework, say so and map your levels to it explicitly.
The second is what it costs, in full. Tuition, materials, registration, exam fees, and what happens if a student pauses. Schools hide pricing behind a contact form and lose the recommendation entirely, because an assistant will not send someone to a page that refuses to say the price when three competitors publish theirs. If your pricing varies, publish the range and the variables.
The third is class size and teacher qualification, stated as facts rather than adjectives. Not small classes with qualified teachers. A maximum of nine students, average of six, all instructors holding a recognized teaching certificate, with the specific certification named, and the share who are native speakers of the language they teach.
The fourth is exam preparation, if you do it. Which exams, what the pass rates have been, how many weeks of preparation you recommend, and whether you administer the test or only prepare for it. This is the highest intent question in the whole category, because a person preparing for a required exam has a deadline and a budget.
The fifth is the practical logistics of starting. When terms begin, whether there is rolling enrollment, how placement works, whether there is a trial class, what happens in the first session, and what a student needs to bring. This is the page that converts the nervous adult, and it is also the page assistants reach for when asked how to get started.
The sixth is who your school is not right for. This is the page nobody writes and the one that produces disproportionate trust. If you do not serve absolute beginners well, say so. If your program suits people who can commit to homework and frustrates people who cannot, say that. An assistant asked to compare options gives real weight to a source that disqualifies itself honestly, because that is a signal of reliability it can rarely find.
A note on format, because it changes whether these pages work. Write the answer in the first two sentences of the page and then explain it. Do not build up to it. Assistants and readers both take the opening, and a page that spends four paragraphs on context before answering gets abandoned by one and summarized badly by the other. Question as the heading, answer immediately beneath it, detail after that.
Keep each of the six on its own page rather than combining them into one long guide. A single page trying to answer six different questions gets retrieved for none of them cleanly, because the relevant passage is buried among five irrelevant ones. Six focused pages, each linking to the others, outperform one comprehensive page by a wide margin for this purpose.
Corroboration is what turns a claim into a citation
A school can write six perfect pages and still not get named, because self-published claims carry limited weight on their own.
Assistants assemble answers from multiple sources and prefer facts that appear in more than one place, particularly when at least one of those places is not owned by the subject. This is the part of AEO for language schools that looks like public relations, because it is.
What counts as corroboration is broader than most schools assume. Coverage in a local outlet. A quote in a piece about workforce language requirements. A listing in a professional body’s directory. An accreditation page on the accreditor’s own domain. A mention in a piece about your city’s changing demographics. Each of these gives the system an independent anchor for something you have also said yourself.
The quality bar is that the source is indexed and identifiable. Roughly ninety-six percent of the thousand-plus publications in the Instant Press catalog are indexed by Google, and indexing is the threshold condition for a mention to enter the pool an assistant can retrieve. A placement in an outlet nobody indexes builds nothing for this purpose, whatever it does for a logo wall.
Consistency across those sources matters as much as their number. If your school is described as offering six languages on your own site, four in one directory, and eight in an old article, the system has a reliability problem and tends to resolve it by naming someone else. Audit every external listing you can find and make them agree. This is dull work with a genuinely high return, and it takes an afternoon.
Reviews sit in an odd position here and are worth understanding properly. Assistants read them, but they use them differently from the way a person does. A person scans the star rating. A system reads the text and extracts claims from it, which means a detailed three-star review mentioning class size and teacher names contributes more usable material than twenty five-star reviews saying great school. Encouraging specific reviews rather than merely positive ones is the higher-value ask, and it is a request students are happy to honor when you explain why.
The last piece is entity clarity. Your school needs to be recognizable as one thing with one name, one location, and one description, appearing consistently across your site, your listings, and your coverage. Schools that have rebranded, moved, or run multiple trading names are frequently invisible to assistants for this reason alone, and the fix is mechanical rather than creative.
It is worth saying what does not work, since the category is being sold a lot of it. Stuffing a page with question-shaped headings you do not answer achieves nothing. Adding structured data markup to a page whose content is vague does not make the content usable. Publishing twenty thin pages about language learning in general, with no connection to your school or your city, builds no authority and dilutes what you have. The mechanism rewards substance, and there is no formatting trick that substitutes for it.
One final framing that helps with prioritization. AEO for language schools is not a marketing discipline bolted onto the school. It is the discipline of writing down what you already know and currently only say out loud on the phone. Every answer on that list of six is something your admissions staff explains twenty times a week. The work is transcription, not invention, which is why a small school can do it properly in a fortnight.
What to do this month
Write the level timeline page with real observed numbers, publish the price, and state plainly who the program does not suit. Those three take a week and they are the three an assistant most needs. Then find every external listing of your school and make them match. Then get one local placement that corroborates the pitch. That order matters, because corroboration of vague claims achieves nothing, and specific claims with no corroboration get discounted.