Almost every language school begins the same way. A good teacher acquires more students than they can take, starts turning people away, and works out that the difference between what a school charges and what it pays them would, at some volume, be a business. That instinct is sound and the resulting schools are often excellent. What catches people out is that running a school is a different job from teaching in one, and the parts that decide whether it survives — pricing, cash timing, class size, who you actually serve — are settled in the first few months, usually by accident.
This cluster is about settling them on purpose. It will not tell you what a language school costs, because that depends on a city, a country and a set of choices you have not made yet, and any figure quoted at you by a stranger is worth exactly nothing. What it will do is give you the blocks the estimate is built from, the decisions that determine the answer, and the arithmetic to run on your own numbers.
"A language school" is not one business. A school teaching exam preparation to teenagers whose parents are paying, a school teaching business English to companies who invoice at thirty days, a school teaching arriving migrants who need certification for a visa, and a school teaching retired hobbyists are four different companies with different selling seasons, different price tolerance, different cancellation behaviour and different cash timing. They can occupy identical premises and look the same from the street.
Choose one to start. Not for ever, and not exclusively — but the school has to be designed for somebody, and the somebody determines your timetable (children come after school, adults come at seven, companies want mornings on site), your pricing model, your marketing, and whether you are selling to the learner or to someone else on their behalf. Schools that try to serve everyone from month one end up with a timetable that suits nobody and a message that persuades nobody, and the person paying for that discovery is you.
The single most useful exercise before anything else: write one paragraph describing a specific person who will enrol, including what they want the language for and what happens to them if they get it. If you cannot, you do not yet have a school; you have premises.
The cost of starting divides cleanly. Space: rent, deposit, fit-out, furniture, insurance — or nothing, if you start online. People: teachers, and the fact that they are paid whether or not the class filled. Getting found: a website, signage, local listings, advertising, and the time you spend on it. Running it: software, payments, accounting, bank charges. And working capital: the months you pay for all of that before enough fee income arrives.
Almost everyone underestimates the fifth and it is the one that shuts schools that were otherwise working. Enrolment is seasonal, a first cohort is small, and the gap between opening the doors and reaching the class sizes you modelled is measured in terms rather than weeks. A school that funds three months and needs eight does not fail because the idea was wrong. It fails because it ran out of money while the idea was working. Fund the gap, count it as a startup cost rather than a rainy-day fund, and know the number before you sign a lease.
The fourth block is the one that has genuinely collapsed in price, and it is worth knowing before you budget from someone else's decade-old figures. Administration software for a small school is no longer a capital purchase. Ours is free for one educator and ten students, €29 a month for a school on Pro, and from €99 on Academy, with 0% taken from what students pay you — and the pricing and total cost cluster explains how to compare that honestly against anyone else, including where competitors are cheaper.
Whether you need a licence to call yourself a school, whether accreditation is optional or effectively mandatory in your market, whether you may sponsor student visas, what company form is sensible, and how tuition is treated for tax all differ by country and sometimes by region. Nobody writing on the internet can answer them for you, and an article that pretends otherwise is doing you harm. Anyone who tells you the rules are broadly the same everywhere has not opened a school in two countries.
What is general is the sequencing advice: find out before you commit to premises or to a market. Accreditation in particular can carry requirements about room sizes, teacher qualifications, published complaints procedures and record retention — all of which are cheap to design in at the start and expensive to retrofit. If any part of your plan depends on students obtaining visas, treat that as the first thing to confirm rather than the last, because it can invalidate the whole model, and it has a way of being discovered in month nine.
Premises give you presence, walk-in enrolment, a community that keeps students through the difficult middle months of learning, and the ability to teach children whose parents want them somewhere supervised. They also give you the largest fixed cost you will carry, committed before you have a single student, and a catchment limited to how far someone will travel on a Tuesday evening.
Online removes the fixed cost and the catchment limit, which sounds strictly better and is not. You compete with everyone rather than with the school down the road, retention is harder because leaving costs the student nothing, and you will spend on acquisition what you saved on rent. The honest summary is that online lets you start smaller and prove demand; premises let you defend a local market once you have. The third article works through both properly, including the hybrid arrangement most schools drift into and why drifting into it is worse than choosing it.
A first term needs surprisingly little: a way for someone to find you and enquire, a placement conversation that puts students at roughly the right level, a timetable of a few classes at times your chosen students can attend, a way to take money reliably, a register, and a person who answers messages the same day. Everything else — the second site, the exam centre, the corporate arm, the app — is a later decision that will be better made once you know who actually enrolled.
What that short list needs is not more features but more care. The placement conversation decides whether students stay, because a learner in the wrong level leaves within a month and tells people why. Taking money reliably decides whether you can pay teachers in January. And answering the same day is, in the early months, most of your marketing, because a new school has no reputation and responsiveness is the only evidence a stranger has that you are real.
Build the cost estimate from blocks rather than borrowing someone's number, settle the ten decisions a plan exists to settle, and choose online or premises deliberately. Each article ends with a checklist.