A package is supposed to delay the next purchase. Then a dashboard punishes it for doing exactly that.
In Swum's eligible first-product cohorts, 84.2% of family-provider pairs beginning with an individual order had another completed order within 90 days. Plans reached 89.2%. Packages reached 52.5%, and group products 50.0%.
The tempting conclusion is that plans retain and packages fail. The correct conclusion is that repeat purchase measures product timing as well as customer behavior.
| First observed product | Eligible pairs | Providers | Another completed order within 90 days |
|---|---|---|---|
| Individual | 467 | 33 | 84.2% |
| Plan | 278 | 16 | 89.2% |
| Package | 118 | 13 | 52.5% |
| Group | 94 | 8 | 50.0% |
Why the package clock runs differently
Suppose a family buys eight lessons and uses one per week. It can remain active for two months without another transaction. Add vacation, cancellation credits or biweekly cadence, and the next purchase may fall outside 90 days.
A plan may produce recurring charges inside the same window. An individual buyer must purchase repeatedly to continue. The transaction rhythm is built into the product.
Comparing their repeat-purchase rates without that context is like comparing a grocery store with an annual membership business by checkout frequency.
Measure the promise of each product
Individual lessons
Measure time to next booking, second-lesson continuation and conversion into a structured product.
Packages
Measure credit utilization, time between lessons, expiration, completion of the bundle and repurchase near depletion.
Plans
Measure successful renewal, lessons delivered per billing cycle, pauses, cancellations and continuation after the initial term.
Group products
Measure enrollment-to-attendance, series completion, re-enrollment into the next session and progression to the next level.
The correct denominator follows the customer promise.
A lower purchase rate can conceal healthy activity
Our separate lesson-continuation analysis found that 85.5% of eligible family-provider pairs had another observed lesson within 30 days of their first observed lesson. That result spans formats and should not be directly substituted for package retention, but it demonstrates why activity and purchase must be kept distinct.
A family can be swimming without buying. A family can buy without swimming. A useful retention system sees both.
The product ladder question
Operators should also ask what each product is meant to become.
- Does a first individual lesson lead to a package?
- Does a package lead to a recurring plan?
- Does a group series lead to the next level?
- Does a plan preserve enough flexibility to survive vacations and disruptions?
Track transitions, not only isolated products. The strongest model may not win every 90-day metric; it may move the right family into the right next commitment.
What this analysis cannot establish
The cohorts are observational and differ in size and provider mix. “First observed” is limited by available platform history. A subsequent order is not necessarily renewal, and no adjustment was made for package size, plan billing frequency, season or customer goals.
The data does not crown a winner. It exposes a measurement trap.
If a product includes more future value in one purchase, expect the next purchase to arrive later. Judge it by the value it promised to deliver.
FAQ
Which first product had the highest 90-day repeat-purchase rate?
Plans had the highest observed rate at 89.2%, followed by individual orders at 84.2%.
Why were package and group rates lower?
One package or group-series purchase can cover multiple future lessons, delaying the need for another transaction. The result should not be read as churn by itself.
What should package retention measure?
Track credit use, lesson activity, bundle completion, expiration, repurchase near depletion and continued family-provider activity.
Review the aggregate product cohorts and choose a model with the packages vs. plans vs. drop-in guide.
