Kestrel Fitness source check and data gap notice, October pack
Kestrel Fitness source check and data gap notice, October pack
Owner: Arjun Mehta. Updated Oct 1, 2026.
Source check results
Finance export failed the pre-release check: the cancellation section is held
The finance export for the October pack arrived with 30 percent fewer rows than the previous period, while the membership data looked normal. The export may be a partial load, so the cancellation figures cannot be verified yet. Northfold has held the cancellation section in draft. The rest of the pack goes out on time on Mon, Oct 5, 2026, and the held section says why it is missing. No unverified figure will be released.
A notice to the Kestrel data owner before release: which source failed the check, which sections are held, and when the held figures will be released. Prepared by Arjun Mehta, analytics lead at Northfold Analytics, for Grace Whitfield, head of membership data at Kestrel Fitness. Notice issued October 1, 2026.
Northfold checks each source against the thresholds below before a section is assembled. Two of the three sources passed every check that could be run.
| Source | Check | Threshold | Result | Status |
|---|---|---|---|---|
| Membership system | Arrival | By 06:00 on Oct 1 | Arrived 05:12 | |
| Membership system | Row count against the previous period | Within 5 percent | 1.9 percent higher | |
| Club check-ins | Arrival from all 14 clubs | By 06:00 on Oct 1 | 14 of 14 arrived by 05:40 | |
| Club check-ins | Row count against the previous period | Within 5 percent | 2.7 percent lower | |
| Finance export | Arrival | By 06:00 on Oct 1 | Arrived 22:10 on Sep 30 | |
| Finance export | Row count against the previous period | Within 5 percent; hold above 10 percent | 28,840 rows against 41,200, which is 30 percent lower | |
| Finance export | Cancellations reconcile to membership status changes | Within 1 percent | Cannot run until a full export arrives |
Row counts and arrival times are sample figures. The export arrived on time, which is why the arrival check passed even though the file was incomplete.
What is held and what is released
The held section appears in the pack as a page that states the reason and the planned release date, so Grace reads about the gap in the pack itself.
How the check decides
| Check | Result | What happens |
|---|---|---|
| Did the source arrive by 06:00? | No | Hold the section it feeds and tell the client |
| Row count within 5 percent of the last period | Yes | Release the section |
| Row count 5 to 10 percent away | Caveat | Release with a caveat printed on the page |
| Row count more than 10 percent away | Hold | Ask the Kestrel data team for a full export |
| Full export confirmed | Yes or no | Release the section, or keep it in draft and report status |
Working thresholds for this retainer (sample figures)
Within 5 percent of the previous period passes. Between 5 and 10 percent is released with a caveat printed on the page. Above 10 percent holds the section. Arjun Mehta and Grace Whitfield can change these numbers, and a change is recorded here before it is used. The finance export is 30 percent lower, so it falls in the hold band.
Release plan
The date that matters is the full export, and it sits with the Kestrel data team. If a full export is not confirmed by Oct 8, the section stays in draft, Arjun Mehta reports the status to Grace Whitfield at the weekly check-in, and the pack keeps stating the reason. The cancellation figures reach Kestrel only after Sofia Reyes has reconciled them to membership status changes.
What this notice does not cover
This is Northfold's pre-release review of the three sources named above. It does not replace Kestrel's own monitoring of its data pipelines, and it cannot say why the export is short. Finding the cause is for the Kestrel data team.
Contacts
Demo client portal. Northfold Analytics, Kestrel Fitness, the people and every figure are fictional sample data. Built on FuseBase.