A regular visitor misses a week, a weekly visitor comes on their usual Saturday, and a frozen member does not attend. A single seven-day absence list can put all three into the same queue even though they need different decisions.
Attendance decay is a change from a member’s own established visit pattern. Check the records and membership context before treating that change as a reason to contact them. A missing visit is not a cancellation, a diagnosis or proof of dissatisfaction.
This guide provides a manual review worksheet, not a claim that Gym Ledger automatically predicts churn. Use valid attendance records alongside membership status and a restricted exception log. If an input is missing, mark the result unverified rather than making the missing information look like a zero.
Build a comparable baseline
Choose complete, comparable weeks before the review week. Record the dates, facility timezone and definition of a visit. For the example below, one member attending twice on the same local calendar day counts as one visit day. That prevents repeated entry scans from inflating the baseline.
Four prior complete weeks are used here to make the arithmetic easy to inspect. Four is a local example, not a validated universal threshold. A new member with only one week of history needs an onboarding check, not a confident decline score.
| Input | Example rule | What to check before accepting it |
|---|---|---|
| Baseline period | Four complete weeks before the current week | Do not include the current week in its own baseline |
| Baseline frequency | Total valid visit days divided by four | Remove duplicate scans under the written rule |
| Current frequency | Valid visit days in one complete week | Do not compare Monday morning with four full weeks |
| Membership eligibility | Membership has started and is not paused for the review period | Keep pauses and delayed starts separate |
| Operational exceptions | Closures, missing device data or rejected access | Resolve collection failures before calling them absence |
If the gym was closed for several days or attendance collection failed, do not automatically scale a short period up to a normal week. Keep it out of the comparison or label a separately defined adjusted calculation. The simplest honest report may be “not comparable this week.”
Work through four different members
These records are fictional. The worksheet is not an industry benchmark and the member labels do not identify real customers.
| Member | Previous four weekly visit-day counts | Baseline per week | Current week | Review decision |
|---|---|---|---|---|
| DEMO-A, frequent visitor | 3, 3, 3, 3 | 12 ÷ 4 = 3 | 1 valid visit day | A real change worth checking after exceptions |
| DEMO-B, weekly visitor | 1, 1, 1, 1 | 4 ÷ 4 = 1 | 1 valid visit day | No weekly decline; do not flag just because six days passed |
| DEMO-C, paused member | 2, 2, 2, 2 | 8 ÷ 4 = 2 | 0 during an agreed freeze | Not comparable; follow the pause and return agreement |
| DEMO-D, access problem | 3, 3, 3, 3 | 12 ÷ 4 = 3 | 0 valid visits, 2 rejected attempts | Access-service investigation before retention outreach |
For DEMO-A, the absolute change is 1 − 3 = −2 visit days per week. The percentage decline is (3 − 1) ÷ 3 × 100 = 66.7%, rounded to one decimal place. That describes a change in recorded behaviour, not a 66.7% probability of cancelling.
For DEMO-B, the change is zero. A message implying they stopped attending would be wrong. For DEMO-C, displaying a 100% decline without the freeze context would be misleading. For DEMO-D, the gym has evidence of attempted attendance; a motivational message would ignore the service problem.
When the baseline is zero, percentage decline is undefined. Write “insufficient baseline” and review the start date and onboarding records. Do not replace the denominator with one just to make the spreadsheet calculate a number.
Use review windows as queues, not verdicts
Elapsed time can help staff organize work, but the same duration does not mean the same thing for every member. Use the following example as a review agenda rather than a rule to contact everyone automatically.
| Review point | Question | Next action | Stop condition |
|---|---|---|---|
| Seven days | Did someone with a frequent routine miss expected visits? | Check complete records, status and access exceptions | No comparable history or an explained pause |
| Fourteen days | Is a changed pattern continuing across another complete week? | Review previous contact before assigning one service check | Member declined contact or another staff member already owns the issue |
| Thirty days | Is membership still active, paused, expired or cancelled? | Route to the appropriate membership or return process | Do not keep someone in an active-member queue after their state changed |
Do not infer the reason from an attendance chart. Work schedules, travel, the gym’s opening hours, a change in routine and service problems can all require different conversations. Staff need only the practical action, not private details that the member has not chosen to share.
Complete a review record before contacting anyone
Here is a filled example for DEMO-D:
| Field | Completed review |
|---|---|
| Review date and window | Monday opening, previous complete Monday-to-Sunday week |
| Observed signal | Baseline 3 visit days; current 0; two rejected entry attempts |
| Membership check | Current plan active in the reviewed period |
| Data check | Other valid attendance was recorded on both attempt dates |
| Working explanation | Entry problem requires investigation; cause not yet confirmed |
| Responsible person | Duty manager |
| Next action | Inspect access configuration and arrange a member-approved entry check |
| Due time | Before the member’s next agreed visit |
| Closure evidence | Entry verified through the approved process; member informed |
Do not add a successful attendance event merely to clear the exception. If the person did not enter, retain the distinction between an attempted visit and a completed visit. Correct erroneous records through the authorized process, with the reason visible to the people reviewing the report.
Make one useful contact, not a pressure sequence
For a permitted service check, staff could say: “Hi, this is the reception team. Is there anything practical we can help with before your next visit?” For an access issue, make the message specific: “We are checking an entry issue. When would it be convenient to help you test access?”
Use the member’s agreed channel and respect contact restrictions. Avoid exposing attendance history in a message preview or asking for sensitive explanations. A person who says they do not want follow-up should not receive another sequence because the same signal appears next week.
Record whether the member responded, whether help was requested, the agreed action and any stop instruction. “No reply” is not permission to keep escalating contact frequency. Assign one owner so reception and a trainer do not independently send the same check-in.
Review whether the process is useful
Count reviewed signals, data errors, explained pauses, access problems, permitted contacts and completed support actions separately. A list of ten flagged people might contain three data problems and two known pauses; calling all ten disengaged members would overstate the result.
Track later attendance using a fixed follow-up window, but remember that a return after contact is not automatic causal proof that the message caused it. Compare similar periods and inspect the underlying cases before changing the threshold. Do not use a small synthetic example to promise a retention improvement.
Use the retention and churn worksheet for membership outcomes, the weekly owner scorecard for review decisions, and the member feedback guide for resolving the issues people actually report.
Frequently asked questions
How do you calculate attendance decline?
Compare valid visits in a complete current period with the member’s average across comparable prior periods. Percentage decline is baseline minus current frequency, divided by baseline, multiplied by 100. A zero baseline cannot produce a meaningful percentage.
Should every member absent for fourteen days be contacted?
No. Check their normal routine, membership state, freezes, operational exceptions, previous contact and communication restrictions first. A time window is a review prompt, not evidence that every person needs the same message.
Does declining attendance prove a member will cancel?
No. It describes a change in recorded visits. Investigate practical obstacles and agreed pauses, keep uncertainty visible, and measure actual membership outcomes separately.
