A booking is a transaction. A lesson is delivery. Confuse them, and a recurring business can look smaller precisely when it is becoming stronger.
In Swum's benchmark, individual lessons represented 58.8% of completed parent bookings but 46.1% of countable lesson records. Plans moved in the opposite direction: 23.9% of completed bookings and 40.6% of lesson records.
The cohorts are not identical, so those gaps should not be treated as a formal conversion ratio. They still illustrate an essential fact: one purchasing decision can create many future service events.
The transaction-count trap
Imagine two operators.
Operator A sells ten individual lessons. The report shows ten bookings and ten lessons.
Operator B sells two recurring plans that each generate eight lessons. The report shows two initial purchasing decisions and sixteen lessons to deliver.
If you rank performance by booking count, Operator A appears five times busier. If you rank it by delivered lesson inventory, Operator B is larger. If you look only at collected cash, billing timing may tell a third story.
None of the reports is wrong. The question attached to the report is wrong.
Four numbers every recurring operator needs
1. Purchasing decisions
How many successful orders, enrollments or renewals occurred? This measures commerce events.
2. Lessons created
How much future instructional inventory did those decisions generate? This measures obligation and capacity.
3. Lessons completed
How much instruction was actually delivered? This is closer to realized service volume.
4. Active family-provider relationships
How many ongoing relationships exist, regardless of how many documents or transactions represent them? This is the human unit underneath the database.
Why plans expand the denominator
A recurring plan does more than schedule future dates. It moves decisions from repeated acquisition to continuation. The family is no longer asking “Should I buy Tuesday?” every week. The operating system asks “Is anything changing about Tuesday?”
That can improve calendar stability, but it also creates obligations: blocked-date handling, instructor substitutions, credits, pauses and clear renewal communication. Recurrence is not free revenue. It is revenue earned through reliable future delivery.
Packages create a different illusion
A package can produce several lessons without another purchase. A 90-day repeat-purchase rate may therefore look lower for package buyers even when those families are actively swimming. The product is doing what it promised: delaying the next transaction.
That is why retention should be measured through both activity and purchase. A customer can be retained without purchasing again this month, and can purchase again without completing much instruction.
Build a metric ladder
For each product, connect the stages:
Order → entitlement → scheduled lesson → completed lesson → continued relationship → next purchase
Inspect drop-off between adjacent stages rather than compressing them into one “booking” number. When a plan underperforms, you will know whether the issue is checkout, schedule creation, attendance, continuation or renewal.
The lesson is not that bookings are unimportant. It is that the value of a booking depends on what it sets in motion.
FAQ
Why do plan lessons have a larger share of lesson records than bookings?
One plan purchase can generate multiple lesson records over time, so lesson volume can exceed the number of initial purchasing decisions.
Is booking count a good measure of swim school growth?
It is one useful measure, but recurring operators should pair it with active relationships, lessons created, lessons completed, revenue and continuation.
Is repeat purchase the same as retention?
No. A package customer may remain active for weeks without purchasing again, while a new purchase does not guarantee future lesson completion.
Explore the full platform benchmark and compare packages, plans and drop-ins.
