A Class That Never Filled: When to Run It, Merge It, or Close It

Why does a half-empty class survive a whole term?

Two children turned up for the Tuesday beginner slot, and the instructor taught it anyway, because nobody had decided in advance what two children meant. The class never reached four, it ran for eleven more weeks, and the loss was already spent by the time anyone looked at the timetable again.

The fix is almost entirely made before the term opens, in two parts. One is the minimum size for each class on the timetable, and the other is the date by which someone has to act on it.

The number you should have set before enrollment opened

Every class on your timetable has a point below which running it costs you money, and the arithmetic is short enough to do on the back of the enrollment sheet. Add what one session of that class costs you to put on, then divide it by what one student pays for one session.

What goes into the cost is the instructor's fee for the session, plus the share of rent, electricity and front-desk time that the hour consumes. Where an instructor is paid per session the first figure is already known, and where they are salaried you are dividing their monthly pay by the sessions they actually teach.

The answer is the break-even head count for that class, and the minimum you publish should sit one student above it. One above, because the class that opens exactly at break-even becomes a loss the first time somebody withdraws, and somebody always withdraws.

Run the sum per class rather than once for the centre. A Saturday group taught by your most expensive instructor in your largest room does not break even at the same number as a weekday slot with a junior tentor.

When do you decide, and what do you tell parents?

Put a decision date on the timetable before you take the first booking, set about a week before the class is due to start. Say on the same line that a class needs a minimum number to run, and that families will hear either way on that date.

Stating it up front is what makes the later conversation survivable. A parent told in advance that a class may not go ahead treats a cancellation as a published rule; the same parent told in week two treats it as the centre changing its mind.

If enrollment is still short on the decision date, give it one week of term and no more. One week tells you whether the two who booked will actually attend and whether a late enquiry converts, and it is short enough that nobody has rearranged their month around a class you then close.

Make one person responsible for running that check on the date. A rule with no name attached to it is the rule that quietly does not happen in a busy enrollment week.

Merge, move, or close: which one fits

Three moves cover almost every thin class, and the one you pick depends on what is actually short, not on which is easiest to explain. Work through them in that order.

Merge the class into an adjacent one where the level and age band genuinely overlap and the receiving class stays under its cap. The test is whether the instructor can teach both groups in one room without the stronger students idling, and for a two-level gap or a mixed age span of more than about three years the honest answer is usually no.

Move the slot where the enquiries came in but the hour did not work. A class that drew interest and lost it at the point of booking is a timetable problem, and shifting it to a later weekday evening or a weekend morning is cheaper than relaunching it next term.

Close it where neither holds, and do that on the decision date rather than at the end of the month. A class with no overlapping level and no better slot will not recover inside the term, and every week it runs takes instructor hours and room capacity from something that would have filled.

Write down which of the three you chose and why. Next term you will be deciding about the same slot again, and the record of the last decision is the only thing that makes the next one faster.

Running it anyway, on purpose

Sometimes the right answer is to teach the thin class, and that is defensible when it is costed rather than drifted into. Three cases usually justify it.

The first is a terminal or advanced level with nowhere else for those students to go, where closing it pushes a family who has been with you for years to a competitor. The second is a salaried instructor who would otherwise be idle in that hour, where the marginal cost of running is close to the room and the lights.

The third is a feeder, where the level exists to deliver students into a class above it that does fill. A beginner group of three that reliably sends two students up a level is buying next year's enrollment, and judging it on this term's contribution alone is the wrong frame.

What separates the deliberate version from the drifted one is that you know the figure. Decide what the term will cost you, write it down, and set a date to look again, because a class carried on purpose for two terms is almost always a class nobody reviewed.

What do the families who booked early get?

The two who signed up first did what you wanted them to do, and a cancellation hands them the worst outcome of anyone on your timetable. They lose the slot they planned around, and they are the ones who will tell other parents how it was handled.

Give them the choice before you announce anything, and give it to them by phone rather than in a broadcast. Offer the place in the class you are merging into, the moved slot if there is one, or a repriced shorter block of private or semi-private sessions, and let them pick.

Where none of that works, return the money rather than issuing a credit by default. A credit keeps the cash and loses the family, and the parent who was refunded cleanly is the one who books again next term.

Hold the fee until the decision date if you can. Taking the full term's payment for a class you are about to close, then reversing it, costs more goodwill than the delay in collecting ever does.

The slot was wrong, not the class

Before you relaunch anything, work out which half of the offer failed, because the two look identical in an enrollment count of two. The question is whether families asked about the class and did not book, or whether nobody asked at all.

Enquiries that arrived and did not convert point at the hour, the price or the trial. Compare that class's trial-to-enrollment ratio with your classes that filled, and check whether something else on your own timetable sat at the same time and took the same students.

No enquiries at all points at the level, the age band or the demand. A beginner group for an age you have never taught is a different bet from a slot that failed, and the cheap way to test it is a short holiday course or a one-off taster rather than a term-long class on the published timetable.

The occupancy numbers you need for this are the same ones worth tracking all term, which is a separate job covered in class occupancy rate by weekday slot. The opposite problem, a class that has outgrown its room, is covered in when to split a full class.

Where does software help here, and where does it not?

The part a system genuinely takes off your desk is knowing, on the decision date, exactly who said yes and who is waiting. Happy Kamper's scheduling and enrollment holds the timetable, the enrollment records and a waitlist per class, so the count you are deciding on is the real one rather than a figure reconstructed from chat messages.

What it does not do is tell you the class is losing money. It does not hold your instructor's cost per session, does not compute a break-even head count, does not watch a date for you, and does not detect that two of your own classes are competing for the same students at the same hour.

Those numbers come from your own costing, and the provider ROI calculator is where to work them out. For bimbel and language centres the vertical view sits on the bimbel and language classes page.

No platform decides whether to close a class. It can only make sure the number in front of you on the decision date is accurate, and that the decision date was not the thing you forgot.