Finding 1
Weekly frequency overwhelmingly dominated
448 of 473 plans with measurable frequency scheduled one session per week. Twenty-three scheduled two sessions and two scheduled three.
The result describes how plans were configured on Swum. It does not determine the ideal frequency for skill development or safety outcomes.
Finding 2
Flexible scheduling was the default pattern
431 plans used a flexible schedule and 42 used a fixed schedule. Flexibility can help families accommodate changing calendars, while fixed slots can simplify capacity planning.
A future outcome analysis should compare rescheduling, attendance and cancellation—not assume either setting is superior from configuration counts alone.
Finding 3
The typical observed canceled plan lasted about three months
Among 194 canceled plans with a usable created and effective cancellation date, the median observed tenure was 90.8 days. The 25th percentile was 42.7 days and the 75th percentile was 163.9 days.
Active plans had a median observed age of 78.1 days at the study boundary. Active-plan age is censored because those relationships had not yet ended.
Finding 4
Semi-private plans represented a substantial minority
123 of the 474 plan records were marked semi-private. That structure can combine recurring revenue with more than one swimmer, but the present study does not normalize price or instructor capacity per participant.
The median plan contained two renewal-history entries, while the 75th percentile contained six and the 90th percentile contained 10.
Plan configuration
| Measure | Records | Share |
|---|---|---|
| One session per week | 448 of 473 | 94.7% |
| Flexible schedule | 431 of 474 | 90.9% |
| Fixed schedule | 42 of 474 | 8.9% |
| Semi-private | 123 of 474 | 25.9% |
| Charge per lesson | 8 of 474 | 1.7% |
Observed canceled-plan tenure
| Percentile | Days |
|---|---|
| 25th | 42.7 |
| Median | 90.8 |
| 75th | 163.9 |
| 90th | 243.6 |
Based on 194 canceled plans with valid start and effective cancellation dates.
Methodology
How the benchmark was produced
Swum aggregated configuration and lifecycle fields from 474 production subscription-plan records without reading customer names, payment credentials or lesson objectives.
- Status is a snapshot at the August 1, 2026 observation boundary.
- Frequency used the structured sessions-per-week field where present.
- Canceled tenure required both a valid created date and effective cancellation date.
- Active-plan age is right-censored and should not be compared directly with completed tenure.
Operator response
What to do with this finding
- Use weekly as a tested starting structure while allowing your own demand and teaching model to decide frequency.
- Review plan health before the three-month mark, where the median observed canceled tenure sits.
- Separate fixed and flexible cohorts when measuring cancellations and rescheduling.
- Track completed attendance and realized revenue alongside plan status.
Limitations
What this data cannot establish
- Plan records are not unique legal customers or businesses.
- Renewal-history entries reflect stored operational events and may vary across legacy workflows.
- Configuration counts do not establish instructional or financial superiority.