Collected once, in writing,
before the timetable is planned.
A teacher asked "can you do Thursday at six?" for every new class will, within a month, have a diary nobody else can read and a set of verbal commitments nobody can reconcile. The alternative costs one form per term: a grid of the week, in which each teacher marks the hours they can teach, the hours they would rather not, and the hours they cannot, along with their maximum contact hours and the levels they are qualified and willing to teach. Collect it three weeks before you plan the term and treat it as a commitment for the term, not a preference to be revisited weekly.
Two details make the grid useful rather than decorative. Distinguish "cannot" from "would rather not", because a timetable that respects every preference cannot be built and a timetable that ignores every one produces resignations; you need to know which hours are negotiable. And record levels honestly: a teacher who can teach an A1 kids class and a C1 exam class is rarer than a teacher who says they can, and the difference shows up in the exam results.
Hours are the number you count.
Preparations are the number that matters.
Schools count teacher load in contact hours because contact hours are what they pay for. Teachers experience load in preparations: the number of different things they have to plan, resource and mark in a week. A teacher with twenty hours across three levels prepares three lessons and teaches each several times, getting better at them; a teacher with twenty hours across seven levels prepares seven lessons, teaches each once, and marks seven different sets of work. The hours are identical. The jobs are not.
When you assign teachers to slots, watch the second number. The timetable that keeps every teacher under four preparations is worth a less tidy grid, because the teacher who stays is worth more than the slot that was slightly better filled. Where a teacher must cover many levels — a small school has no choice — reduce the preparation cost another way: shared materials by level, curriculum that is written down rather than reinvented, and AI-generated lesson plans and activities as the starting point rather than a blank page. On SprintUp that generation is unlimited on the relax lane for every teacher in the school, which changes what seven preparations costs.
Everyone wants four to seven.
Decide who gets it.
The after-school window is where a language school earns most of its revenue, and it is three hours long, four days a week, plus Saturday morning. There are more classes that want those hours than there are rooms and teachers to run them, and the hours a teacher works in that window decide their income. This is a fairness problem before it is a scheduling one, and schools that leave it to chance produce a silent hierarchy in which the teachers who were there first hold the evenings and the newer ones hold the empty afternoons.
Decide the rule and say it out loud. Some schools allocate peak hours by seniority; some rotate them by term; some tie them to the level a teacher is qualified for, because the teens' exam classes that fill the evenings need particular teachers. Any of these is defensible. What is not defensible is the rule that nobody has stated, applied by the owner from memory, which every teacher can see and nobody can question.
Most language teachers are part-time.
Plan as if it is true.
The typical language school teacher has another job, another school, a family or a degree, and teaches you eight to fifteen hours a week in the windows those things allow. The timetable that assumes full-time availability will fail in the second week. Plan in blocks that suit part-time lives: two consecutive evening classes on the same day rather than one class on each of four days, so that a teacher travels once and earns twice; a Saturday morning of three classes rather than a Saturday morning of one. Blocks reduce the number of teachers you need to staff a week, reduce cover complexity, and make the job worth having.
Split days — a morning class and an evening class with nothing between — are the least attractive shape you can offer and should be paid or avoided. A teacher will take one split day for a good school. They will not take three.
Absence is certain.
Only the date is unknown.
Over a term, someone will be ill, someone will have an interview, and someone will not turn up. The cover plan is made when the availability grids are collected, because that is when you know who could teach what and when. It is a list per level: the teachers who could cover an A1 kids class on a Tuesday evening, in order of preference, with their contact details and the hours they have already said they could work. Keep it where the person who takes the sick call at eight in the morning can see it, and keep it current, because a cover list from September is fiction by February.
Decide in advance which classes are covered and which are postponed. A general-English adult class can be covered by any competent teacher with the shared materials for that level; an exam class in the fortnight before the exam should be postponed and added at the end of term rather than handed to someone who does not know the students. Tell families the same morning through whichever channel already reaches them, and record the postponement in the same place the timetable lives so that the end-of-term date moves visibly for everyone.
For teachers, the test is one question:
can I see my week without asking?
Teachers do not want a scheduling system. They want to open something on Sunday evening and see their week: which classes, which levels, which rooms or links, and who is in them, with the changes since last week already there. That is the whole requirement, and it is met by very few spreadsheets and by any platform in which a class is a record with a teacher, a level and a roster that the teacher can see from their own side.
The honest list matters as much as the feature list. No platform solves your timetable; the constraints, the load, the fairness rule and the cover list are decisions, and they are yours. What the software does is make the decision, once made, visible to everyone it affects, and keep it that way when it changes. That is the difference between a timetable that costs you Sunday evening and one that does not.