You probably have a feeling for it. "A few." "Not many." "Some of the Year 6s, but they always go." Sit down with two lists and the feeling turns into a number, and the number is nearly always bigger than the feeling.
The two lists
Get the enrolment list from the last week of Term 4. Then get the list from week 3 of Term 1. Not week 1, because some families come back late, and not the January enrolments, because those are new families and this exercise is about the ones you already had.
Put the two side by side. Every name on the Term 4 list that isn't on the Term 1 list goes on a third list. That's the raw count.
Now take names off it. Anyone who moved house or changed school across town. Anyone who finished Year 12, or sat the selective test they came to you for and was always going to leave afterwards. Those aren't losses, they're the end of the arrangement, and counting them muddies the number you actually need.
What's left is the list of families who could have stayed and didn't. That's your number. Write it down before you do anything else with it.
Why owners guess low
Three reasons, and none of them is carelessness.
The first is that leaving is quiet. A family that's unhappy tells you. A family that's drifting doesn't tell anyone. The form doesn't come back, the January email goes unanswered, and by February you've stopped noticing the gap because someone new is sitting in it. You remember the two who complained. You don't remember the nine who just didn't turn up.
The second is that January fills the room. Enquiries run hot before Term 1, and if the classes look full in week 3, the feeling is that things went fine. But full isn't the same as retained. A class of twelve with five new faces is a class that lost five, paid to replace them, and starts the year with five families who don't know you yet.
The third is that nobody counts. Rosters get rebuilt for the new year, the old list gets archived, and the comparison never happens. Most centres could tell you their enrolments to the student and couldn't tell you their re-enrolments at all.
What the number is worth
Put your own fee against it and multiply. That isn't a scare figure, it's just what those families were already paying you, and it's the same list you'd have taught anyway without a single new enquiry.
The retention cost calculator does this with your own fee and your own count, then shows you how few of those families you'd have needed to hold. That number is the one to remember, because a few retained students pays for most of what you'd do to retain them, including a platform.
The other cost is harder to price. Every seat you refilled took an enquiry, a call, a trial lesson, and a first term where the family was still deciding. A family in its third year doesn't need any of that. They renew on habit, they refer their friends, and they forgive a bad week. Replacing them with a stranger is a step backwards even when the roster looks the same.
What to do with the number
Three things, in order.
First, go back through the third list and write one line next to each name about why. Not a guess about what they were thinking. What you actually know: a call that mentioned money, a parent who never replied to anything, a child who stopped bringing work back in Term 3. You'll find the reasons cluster. Often the largest cluster is silence, families you hadn't heard from in months before they left, and why parents leave in Term 4 is mostly about that cluster.
Second, put the number on the wall as this year's budget. Everything in the Term 4 retention playbook costs time, some of it costs money, and the question every step is measured against is whether it's cheaper than losing the students it saves. Your own number answers that question for most steps before you start.
Third, and this is the one that changes next year's count, start the gone-quiet list in Term 3. Every fortnight, list the students who haven't handed anything back and the parents who haven't replied to anything. That list is next summer's third list, five months early, while there's still time to do something about it.
The version where you don't have to count
On the platform this exercise mostly disappears, because the two lists live in the same place and the gone-quiet list builds itself. The tutor dashboard shows students with no activity for two weeks across every class, so the owner sees the drift in Term 3 rather than discovering it in February.
The bigger change is upstream of that. A parent who has received twenty short reports across the year, each one naming a skill that improved, has a reason to re-enrol that a parent holding four invoices doesn't. The count still matters. It just gets smaller.
This week
Find the two lists. If the Term 4 one is gone, the invoicing system or the class rolls will have it. Do the comparison, take out the movers and the finishers, and write the number down with the fee beside it. Then run it through the calculator and keep the result somewhere you'll see it in Term 3.