A language school is a scheduling and enrolment business that happens to teach. This guide covers the five systems every school runs, the decisions that determine everything after, and where the software fits — for schools online, on premises or both.
This guide is for people who run independent language schools: general English and the other big languages, exam preparation for IELTS, Cambridge, DELF and DELE, after-school programmes for children and teenagers, evening classes for adults, and corporate contracts. It applies whether you are one owner-teacher with a Saturday-morning kids' class or a two-branch academy with twenty teachers on the payroll. It applies whether you teach in a room, online, or both.
Here is the uncomfortable operational truth. Teaching is the part most owners are good at, and it is rarely the reason a school struggles. Schools struggle on everything around the teaching: placing students at the right level, building a timetable that actually fills classes, collecting fees before the term starts rather than chasing them after, showing parents progress they can see, keeping good teachers, and then doing every one of those things again next term. That is why a typical language school runs on five or six disconnected tools and a spreadsheet nobody trusts.
The guide is organised around that reality. Six clusters cover starting, timetabling, online and hybrid delivery, levels and progression, pricing and payments, and growth. Where SprintUp fits, the guide says so plainly, including the plan you would be on and what it costs. Where it does not, it says that too. We built the platform for exactly this kind of school, which is a reason to read carefully, not a reason to skip the comparisons.
Strip away the branding and every language school runs the same five systems. Most owners never name them, which is why they end up buying five products and wiring them together by hand. Naming them is the first step to running the school instead of being run by it.
How a prospective student becomes an enrolled one: the enquiry, the placement test, the level decision, the timetable slot, the payment, the portal login. Every leak in this sequence is a student who paid a competitor instead.
Which teacher teaches which level in which room or video link at which hour, for how many weeks, with what happens when someone is ill. It is the single most-edited artefact in the school and the one most often kept in a tool nobody else can see.
The class itself, and everything that supports it: the video link, the activities, the homework, the attendance register. Online schools live or die here. Premises schools tend to under-invest here and lose students to the ones that did not.
What gets taught at each level, how progress is checked, and how a student is moved up. Schools without a written curriculum depend entirely on whichever teacher is in the room, and feel it the day that teacher leaves.
Pricing, invoicing, deposits, instalments, failed cards, refunds and teacher pay. Fee collection is the system most schools run worst, and it is the one that decides whether the school is still open in three years.
The expensive part of running these as five separate products is not the five subscriptions. It is that the same student exists in five places. A level change made in the teaching tool does not reach the timetable; a timetable change does not reach the invoice; the invoice does not know the student stopped attending in week four. The cost shows up as re-keying, as errors parents notice before you do, and as the owner spending Sunday evening reconciling systems instead of planning the term. Whatever software you choose, the test is whether the student exists once.
New schools spend their energy on the name, the logo and the website. Those are reversible. The decisions below are not, or not cheaply, and they set the cost base, the timetable and the price list before a single student enrols.
Which languages, which levels, which learners. "English for everyone" competes with every school in the city and every app on the phone. "Cambridge exam preparation for 14 to 18 year olds" competes with three schools and can charge for the outcome. Narrow first; widen once the timetable is full.
Online-first, premises, or hybrid. This is the cost decision. Premises means rent, fit-out, insurance and furniture before the first student, and a radius of a few kilometres. Online-first means a much smaller fixed cost and a much larger catchment, and it demands better delivery to keep students. Hybrid is the most common end state and the hardest to start with.
Group size and class length. Groups of six to eight at 60 or 90 minutes are the economic core of most schools: enough students to make the hour pay, few enough that everyone speaks. One-to-one is higher revenue per hour and far harder to timetable. Decide the standard formats now; the timetable is built from them.
Per course, per term, per month, or per package of hours. Each one changes cash flow, churn and the enrolment flow. A school that sells terms collects money before it delivers; a school that sells monthly subscriptions is one bad month from a gap. The Pricing cluster covers this in full.
Where the student exists. Decide this before the first enrolment, because migrating twenty students out of a spreadsheet is annoying and migrating two hundred is a project. The platform you choose should hold the student, the timetable, the classes and the payments in one place, or you will be building integrations instead of a school.
Startup-cost guides for language schools quote wide ranges because the two paths are not comparable. An online-first school can open for the cost of a domain, a laptop and the owner's time: the school website, the live classroom, the student portal and payments can all start on a free plan, and the first paid expense is often a teacher. A school with premises spends on rent, deposit, fit-out, insurance and furniture before it has a timetable. Neither path is wrong. What is wrong is starting on the premises path with online-first economics in the business plan. The Starting a language school cluster itemises both.
A language school timetable is built from constraints, and the school that lists its constraints honestly builds a better timetable than the school with better software. Teachers have availability and a maximum load. Levels need a minimum cohort to run and a maximum to stay a conversation. Rooms and video capacity are finite. Students have the hours they can attend: for children and teenagers that is roughly four to seven on weekdays and Saturday mornings, for working adults it is evenings and lunchtimes, and those windows are the same for every school in your town. Term dates, public holidays and exam seasons move everything.
The practical consequence is that the timetable is not something you fill in; it is something you solve, once per term, and then protect. Every ad-hoc change made during the term to accommodate one student or one teacher costs more than it appears to, because it cascades: the room, the other students, the invoice, the make-up policy. The schools that run smoothly are not the ones with the most flexible timetables. They are the ones with fixed slots per level and a written policy for what happens when someone cannot make it.
Build from the constraints outward. Teacher availability and the after-school and evening windows first; then the level cohorts that fit them; then the rooms or video links. Doing it in the other order produces a timetable that looks full and runs at 60% because the students cannot attend it.
Fix the slot per level per term. "B1 teens is Tuesday and Thursday at five" is something a parent can plan around and a teacher can rely on. Moving it in week three to accommodate one family loses three others.
Write the cancellation and make-up policy before the first booking. How much notice, what counts as a make-up, whether make-ups expire. Announce it at enrolment. The policy you improvise under pressure is always more generous than the one you would have written calmly, and it sets a precedent you cannot take back.
Keep one source of truth. If the timetable lives in the owner's head or a private calendar, every question becomes a message to the owner. Teachers need to see their own classes; students and parents need to see theirs; a change in one place should be a change everywhere.
Review load per teacher every term, not every year. Twenty-two contact hours with three levels is a different job from twenty-two contact hours with seven. The second one produces a resignation in March.
On SprintUp the timetable is held with the classes, the teachers and the students in one campus, so a class has a teacher, a level, a roster and, for online sessions, a live video link generated by the platform rather than a Zoom licence per teacher. The Timetabling cluster goes through building the term timetable slot by slot, the cancellation and make-up policies that survive contact with real families, and scheduling around the teachers you actually have rather than the ones you wish you had.
Online language teaching has a reputation problem it earned in 2020, when a great deal of it consisted of a teacher talking at a grid of muted rectangles. That is not online language teaching; it is television. The schools whose online students finish the course and re-enrol run something closer to the classroom than the classroom itself: groups of four to eight, structured speaking in every session, homework that is set, submitted and marked with feedback rather than checked, and attendance that students and parents can see without asking.
The design question for a hybrid school is not which lessons to move online but which parts of a lesson belong where. Speaking practice, exam-technique sessions and anything that depends on reading the room tend to stay in the room. Grammar input, vocabulary work, listening and self-paced review move online without loss and free the in-person hour for the things that need it. A hybrid course designed this way is more expensive to build and cheaper to run than either pure model, and it is what most schools that start online eventually become.
Above eight, speaking time per student falls below the level that produces progress, and students notice before you do. Six is the number most experienced online teachers settle on.
Breakout pairs with a task and a time limit, rotated every session. "Any questions?" produces silence online in a way it does not in a room.
Set in the platform, submitted in the platform, marked with a comment the student reads. Homework that is merely ticked off teaches students that it does not matter, and they are right.
The tools matter less than the design, but the wrong tools make the right design exhausting. A school needs video with breakout capacity, a place for activities and homework that the same login reaches, and an attendance record that does not depend on the teacher remembering to write it down. On SprintUp, live video classes, activities, quizzes and the student portal are one system with one login per student, which is the difference between a teacher running a class and a teacher running four browser tabs. The Online and hybrid delivery cluster covers designing the course, running the live session, and the tech stack that does not fight you.
The CEFR scale, A1 to C2, is the closest thing language teaching has to a shared unit of measurement, and most schools map their courses to it whether or not they say so on the website. Its value is not precision; it is that a parent in one school, a teacher in another and an examiner in a third mean roughly the same thing by B1. Use it as the spine of the level structure and the placement test, and resist the temptation to invent your own scale for marketing reasons.
Placement is where schools first lose a student's trust, usually by placing too high to make a sale or too low to make a class run. A good placement test is short, covers the four skills unevenly on purpose (a speaking component is not optional, because it is the skill students are paying to improve), and produces a starting point rather than a verdict. The verdict comes from the first two weeks in the class. Tell students that at placement and the first-fortnight moves stop feeling like failures.
Progression is the harder discipline. The easy rule is seat time: twelve weeks at A2, therefore B1. It is easy because it never requires a difficult conversation, and it quietly destroys the school, because within two terms the B1 class contains four levels and the teacher is running a timetable of one. The right rule is evidence at the end of the term: an assessment mapped to the level descriptors, the teacher's judgement, and the homework record. Students who are not ready repeat with a clear reason; students who are ready move. Parents will accept a repeat far more readily when the report shows them why.
That report is the third piece. Parents of children and teenagers want three things from a language school and rarely get them: did my child attend, what level are they, and what is the next step. A progress report that answers those three questions in one screen does more for re-enrolment than any discount. On SprintUp the attendance, the quiz results and the level sit on the same student record, so the report is a view rather than a document someone has to assemble. For the classroom practice underneath progression, the feedback loops and assessment guides go deeper than this pillar can.
Language schools inherit hourly pricing from private tutoring, and it is the wrong model for almost everything except one-to-one. An hourly price invites the student to count hours, to skip the ones that seem optional, and to compare you with the cheapest tutor on the internet. A course price, per term or per level, sells an outcome and a cohort: twelve weeks of B1 with the same six people and the same teacher, ending in an assessment. It also collects the money before the delivery, which for a small school is the difference between planning a term and surviving one.
The four models that work, and who they suit:
The default for group classes. Paid up front or as a deposit plus one instalment. Cash arrives before costs; churn is visible at re-enrolment, where you can act on it.
Suits adult conversation classes and ongoing programmes with no fixed end. Smooth revenue, low commitment, and a cancellation is one click away, so retention is the whole job.
For one-to-one and exam coaching. A block of ten hours at a lower rate than one. Expire them; otherwise you are carrying a liability into next year.
A company pays for a group of employees per term. Higher value, slower to close, invoiced rather than paid at checkout, and the one model where you will genuinely need accounting software.
The enrolment flow is where pricing meets reality. The sequence that converts is short: the programme page on your website, a placement test or a short call, a level and a timetable slot, payment, and a portal login, in that order and without a human having to email a PDF at any stage. Every step that requires the owner to do something by hand is a step where a family goes quiet. Measure the flow where it leaks. Most schools discover the leak is between "we have placed your child at B1 on Tuesdays" and the payment, because that gap is where the parent talks to another school.
Collecting fees is a system, not an event. Deposits at enrolment, a written instalment schedule, automatic retries on failed cards, and a policy on arrears that is applied on day one rather than day sixty. A school that lets a family attend for six weeks without paying has made a donation, and the family knows it.
Nothing. Students pay your school directly through Stripe and SprintUp takes 0% of that payment, on every plan including the free one. Standard Stripe processing fees still apply. Some platforms in this market take a percentage of every payment your students make on top of a subscription; if you are comparing, ask that question first, because at a few hundred students it is the largest number on the invoice. The only place SprintUp takes a share is the curriculum marketplace, where the creator keeps 80%.
Ask any established language school where its students come from and the answer is the same three sources in the same order: word of mouth, local search, and partnerships. New schools tend to spend on the second, hope for the first, and ignore the third. The order of effort should be reversed.
Word of mouth is a system when you make it one. Sibling and family pricing, an end-of-term showcase that parents attend and photograph, a referral that credits both families, and a re-enrolment window that opens before the term ends rather than after. A school that re-enrols eighty percent of its students each term needs very little marketing; one that re-enrols half is marketing to stand still.
Local search is the one channel where a small school can outrank a national brand, because the query has a town in it. "English classes for kids in [your town]" is won by the school whose website has a real page for that programme, with the level, the timetable, the price and the enrolment button on it, on the school's own domain. A single home page with a contact form does not rank, and neither does a listing on a marketplace someone else owns. Programme pages, one per course, are the unit of local search.
Partnerships are the slow channel and the durable one: primary and secondary schools that refer families, employers who pay for staff, universities with international students, relocation agencies. They take a term to open and then produce students every term after.
The second branch comes up once the first timetable is full for two terms running. The mistake is to open it as a second school, with a second website, a second student list and a second set of prices to keep in sync. Open it as a second site of the same school: one account, one student record, separate branding where the town needs it, and a timetable system that already knows both. On Academy a school can run up to five websites on its own domains from one campus; Pro allows three, the free plan one on a SprintUp address. The Marketing and growth cluster covers the programme page that ranks, the referral mechanics, and the second-branch decision in detail.
The software question for a language school is not which product has the longest feature list. It is whether the five systems above share one student record, and what it costs you when they do not. Four questions settle most evaluations. Does the student exist once, or once per tool? When the timetable changes, does the teacher, the student and the invoice all find out, or does someone have to tell them? What happens to your data when you leave: can you export every student, every payment and every attendance record, or are you renting your own school? And where is that data hosted, under which law, because for a school in Ireland or the EU the answer has to be GDPR and the vendor has to be able to say so without checking.
Where SprintUp fits, stated plainly:
One owner-teacher, trying it with a single class: the free plan. One educator, ten students, one website on a SprintUp address, live classes, the student portal and unlimited AI generation on the relax lane. No card. It is a real school, not a trial, and it stays free.
A solo teacher-owner with more than ten students: Pro, at €29 a month or €261 a year. Up to fifty students, three websites with your own branding, the fast AI lane with 5,000 credits a month and premium models.
A school with a team: Academy. Every teacher and every student, priced by the students actually in a running class each month: €99 covers the first 25, then €4 each to 100 and €3 beyond. Five websites, 250 fast-lane credits per active student pooled across the school, and 30 days free for up to 25 students so the first term can start before the first invoice.
On every plan: students pay you through Stripe and SprintUp takes 0% of it; the campus, website builder, live video, student portal, AI curriculum tools and marketplace are included; data is held under GDPR with EU data residency. Domains are sold separately. Prices exclude VAT.
Where it does not fit, also plainly: SprintUp is not a payroll or accounting system, and a school with corporate contracts or more than a handful of teachers will still need one of those alongside it. It is built for schools that want the campus, the classes, the students and the payments in one place, not for schools that want to keep their existing six tools and add a seventh.
If you are comparing, compare properly. The choosing school software guide has the evaluation checklist and the vendor-by-vendor comparisons, and the pricing page has a calculator that shows what Academy would cost for your student numbers before you talk to anyone.
At minimum: a way to enrol and place students, a timetable that teachers and students can both see, a place to run or support classes, somewhere to keep curriculum and assessment, and a way to collect fees. Most schools assemble five or six separate tools for this. An all-in-one school platform covers the same ground with one student record, which removes the re-keying and the errors that come with it.
It depends almost entirely on one decision: online-first or premises. An online-first school can launch for the cost of a domain, a laptop and the owner's time, because the platform, website, live classes and payments can start on a free plan. A school with premises pays rent, fit-out, insurance and furniture before the first student arrives, which typically puts the figure into five figures. The full breakdown is in the Starting a language school cluster.
Yes, and many do. What makes it work is the same as in a room: small groups, structured speaking time in every lesson, homework that gets feedback, and attendance that parents and students can see. What makes it fail is treating a live class as a video to watch. Hybrid schools usually keep speaking-heavy and exam-practice sessions in the room and move grammar, review and self-study online.
Use the CEFR scale (A1 to C2) as the shared language, run a short placement test with a speaking component, and treat the result as a starting point rather than a verdict. Then move students between levels on end-of-term evidence, not on how many weeks they have attended. Progression on evidence is what parents pay for and what keeps a class coherent.
No. Students pay your school directly through Stripe and SprintUp takes 0% of that payment on every plan, including the free plan. Standard Stripe processing fees apply. The only place SprintUp takes a share is the curriculum marketplace, where creators keep 80% of each sale.
Academy: €99 a month covers your first 25 active students, then €4 per active student up to 100 and €3 beyond that, with unlimited teachers. An active student is one enrolled in a running class that month. It starts with 30 days free for up to 25 students, and paying yearly gives you twelve months for the price of ten. Prices exclude VAT; domains are sold separately.
AI lesson plans, exit quizzes, differentiation, and course outlines — all on the free plan.
Set up your school free →See what schools getWhat it costs to start a language school, what to decide first, and how an online-first launch changes the numbers.
How to build a language school timetable that fills classes, handles cancellations and make-ups, and stops eating your evenings.
Running live and hybrid language classes that students complete: platform, video, breakout practice and a tech stack that does not fight you.
Placing students at the right CEFR level, moving them up on evidence, and showing parents progress they can actually see.
How to price language courses, build an enrolment flow that converts, and collect fees without a platform taking a cut.
Getting students for a language school: a website that ranks locally, referrals that compound, and opening a second branch.
Configuring the student portal, designing the enrolment flow, and running the first group through it without losing anyone.
The setup sequence that works: school structure first, then website, classrooms, teachers and students.
What one educator and ten students get for nothing, where the limits are, and what changes on Pro and Academy.