The free class you never agreed to sell

It is October. The messages have started.

A child has had a temperature all week. A family is away at half term. Someone’s car would not start. Each message is reasonable, each one arrives on its own, and each one gets the same answer, because saying yes to a parent you like about one missed class is obviously the right thing to do.

By December you could not say how many free classes you gave away. Not roughly. Not within twenty.

That is not a criticism — it is how almost everyone runs it. The thing worth noticing is that you never made a decision. You made a series of individual kindnesses, each one clearly correct in isolation, and together they became a policy that nobody designed and nobody priced.

Everybody asks this. Nobody has an answer

We read what our customers write to us. Across four weeks this autumn, make-up classes tied with pricing as the most-raised subject — but they came from a far wider spread of businesses. Pricing questions cluster around a handful of accounts setting things up. Make-ups come from everybody, constantly, in every vertical.

And here is the part that made us want to write this. When someone writes to us about make-ups, in nearly two of every five conversations the question is not how to configure anything. It is:

How many should I actually allow?

That is not a software question. It is a business question, and the honest state of the internet is that nobody answers it properly — including us. We have a post on this blog that recommends “one to three per term” and does not show a single line of working. It is reasonable advice. It is also a number we handed over without telling anyone how to check whether it fits their business.

So: here is the working.

A make-up has two prices, and you pay one of them either way

Most writing on this topic picks a side. Either make-ups are revenue leaking out of your business, or they are the warm thing you do for families. Both halves are true at once, and you cannot decide anything useful until you hold them together.

The first price is a seat. Seats are the only thing you sell. Your revenue is seats, times price, times how full you run. A make-up is a seat granted at a price of zero — which makes your make-up policy a discount, handed automatically to whoever asks, in a size nobody ever set.

What that discount costs depends entirely on one thing:

• The class has empty seats. It costs you almost nothing. The instructor is paid, the hall is rented, the lights are on. Say yes, and stop worrying about it.
• The class is full. Now it is expensive. The make-up either displaces a seat you could have sold, or it pushes you past the ratio you promised parents — and in a swim school or a toddler class, that ratio is the product.
• You hold seats back for make-ups. You have already paid for the capacity, so the cost is decided. The only question left is whether you sized it right.

The second price is a renewal you do not get. A family pays for a term, loses three sessions to a winter of illness, is offered nothing, and quietly does not sign up again. They paid full price for less than they bought, and that is the feeling they take into the renewal decision. Research summarised by Harvard Business Review puts the cost of acquiring a new customer at five to twenty-five times the cost of keeping one you already have. Against that, a seat in a class that was half empty anyway is not generosity. It is the cheapest retention spend available to you.

So the question was never should I be generous. It is which of those two prices am I paying, and did I choose it.

Which brings us to the thing we keep seeing.

A provider — and this pattern is common enough that nobody reading it should assume it is about them — was drowning in requests. Instead of publishing a limit, they set the window in which make-ups could be booked to a stretch of dates containing no sessions at all. The booking page still worked. Parents could still click. There was simply never anything there.

Nobody was lied to. Nobody was told no, either. The policy became technically impossible instead of publicly limited, because telling a parent “you get two” feels mean, and having nothing available feels like bad luck.

That is what a policy nobody designed eventually looks like. It is not cynicism, it is exhaustion — and it is entirely avoidable, because the number they needed was sitting in their own register the whole time.

Work out your number

You need three things, and you already have all of them.

1. How full are your popular slots — really? Not your average across everything. The average is flattered by the awkward Tuesday afternoon nobody wants. Take the three or four slots parents actually ask for and work out what share of those seats are sold.

2. How many sessions are in a term, and how many children.

3. Your absence rate. Do not guess this, and do not take ours. Open last term’s attendance and count: total absences, divided by (children × sessions). Most providers are surprised — it is usually higher than they thought and lower than they feared. It is also the single most useful number in this whole exercise, and almost nobody has looked it up.

If your system can be read by an AI assistant, skip the counting

Most of the work above is looking things up, which is exactly the sort of job an AI assistant can do for you if it is connected to the system you already run classes in. If yours is, paste this and let it pull the numbers:

Using my own booking data, help me set a make-up policy. Use the real figures —
do not estimate, and flag anything the data cannot answer.

Clean the inputs first, and tell me what you dropped:
- Skip classes where attendance was never marked. A class showing 0% attendance
  across every session is an unmarked register, not an empty room.
- Skip classes that have already finished, and any with no enrolments — a class
  that has ended often reports zero seats filled even though it ran full.

Then, for the classes that remain:
1. Attendance — what share of booked sessions were actually attended?
2. Occupancy of my busiest programmes — the share of seats sold in the slots
   that fill up, not the average across everything.
3. Make-up demand against available slots, per programme.

Then work out:
- expected make-up requests  = enrolments × sessions × (1 − attendance rate)
- spare capacity              = seats × sessions × (1 − occupancy)
- whether I can say yes to everything, or the make-ups per child I can support
  (spare capacity ÷ children, rounded down)

Finish with the one slot under the most pressure.

Two minutes instead of an evening with a spreadsheet. The last line matters: the slot under most pressure is usually where your policy is quietly costing you sales.

Check the cleaning step before you trust the answer. Two things can quietly ruin the number. A class where nobody ever marked the register comes back as 0% attendance, which reads as a catastrophe rather than a missing habit. And a class that has already finished can report zero seats filled even though it ran full, because the seat count is live and the class is over. Both push the answer in the same direction — towards thinking you have more room than you do. If the assistant does not tell you what it excluded and why, ask it.

Now the arithmetic, which is deliberately crude because precision here is false comfort.

Spare capacity   = seats × sessions × (1 − occupancy of popular slots)
Expected demand  = children × sessions × absence rate

Then one of two things is true.

Spare capacity is comfortably bigger than expected demand. You can afford to say yes to everything — so say so, out loud, on your website. An unlimited policy you have actually costed is a genuine selling point, and it costs you nothing you were going to earn anyway.

Spare capacity is smaller than expected demand. You are already rationing make-ups, whether or not you have admitted it. The only question is whether the rationing is done by a published rule or by whoever emails you first. Your honest number is:

Make-ups per child = spare capacity ÷ number of children   (round down)

Round down, publish it, and stop negotiating case by case.

The question that matters more than the number

Where the make-ups land matters more than how many you grant.

Ten make-ups spread across your quiet slots cost you nothing. Ten that all want Saturday morning cost you Saturday morning — the slot you need free to sell to new families, the one that generates your waiting list, the one your growth depends on.

So whatever number you land on, add the sentence that actually protects you: make-ups are offered in slots with availability, and popular times are not guaranteed. That single line solves more problems than the number does.

The cost nobody puts on the list

There is a third price, and for a small provider it is usually the biggest one.

Several people told us the same thing this autumn, in almost the same words: I end up doing it by hand. Finding a slot with room. Checking the ratio. Emailing the parent. Moving the booking. Chasing the parent who never replied. Doing it again when they cannot make the one you offered.

Time it once — genuinely, with a timer, start to finish — then multiply by how many requests you handled last term. Most owners find the administration costs them more than the seats ever did, and it appears in no budget, no report and no decision.

If the number shocks you, that is not an argument for refusing make-ups. It is an argument for a rule clear enough that parents can self-serve, instead of a negotiation that runs through your inbox every October.

Six lines you can publish this week

Fill these in and put them on your website. It takes ten minutes and it ends the negotiation.

1. You can claim a make-up for ___ missed sessions per term.
2. Tell us before the session, through ___.
3. Make-ups are booked in slots with space. Popular times are not guaranteed.
4. Use it within ___ weeks. Credits do not carry into the next term.
5. Make-ups cannot be swapped for a refund or a discount.
6. If we cancel a session, you get a replacement regardless of the above.

That last line matters more than it looks. The policy limits apply to absences you did not cause. Parents accept limits far more readily when it is obvious the business holds itself to a different standard.

Three numbers to watch

Once the policy is live, these tell you whether it is the right size.

• Make-ups granted per 100 enrolments. Your baseline. Watch the direction, not the number.
• Share claimed versus expired. If most expire, your window is too short or booking is too awkward — and you are getting the resentment without the retention.
• Occupancy of the slots make-ups land in. If they are pushing your best classes to full, the policy is now costing you sales, and the number needs to come down.

One thing we would say as a vendor, and only here

Zooza handles most shapes of this: a cap per term, an expiry window, a waiting list for make-ups, and capacity held back so that catch-up attendees have somewhere to go without your paying customers being squeezed out. If you want the mechanics, they are documented here and in the guide to capacity and extra capacity.

The prompt further up works against any system an assistant can read. If you run on Zooza, connecting it to Claude lets you paste that prompt and get the three numbers back from your own account — there is a report behind it that compares make-up demand against the slots you actually have free.

But the settings are not the point of this piece, and no setting will decide the number for you. The recommendation above is ours, worked out the way we would do it for ourselves — and it is worth more to you than any configuration screen.

If you want the background rather than the arithmetic — what the common policy shapes are and who each suits — we covered that in Make-Up Lessons Done Right. This post is the part that one was missing: how to get your own number instead of borrowing ours.

Sources and method. Customer-conversation observations are drawn from messages our own customers sent us between 12 September and 9 October 2026, counted by theme; they describe patterns across providers in several countries and no individual business is identifiable. Retention economics: The Value of Keeping the Right Customers, Harvard Business Review, October 2014.