The reliability program you actually need.

Most reliability program manuals are written for a fleet size the airline doesn't have.

They borrow structure from major carriers; statistical sampling thresholds, alert level triggers calibrated for dozens of tail numbers, trend analysis built for volume most regional operators will never see. On a small commercial fleet, that borrowed structure doesn't fail loudly. It just quietly stops functioning as a reliability program and becomes a document that exists to be filed.

The small-fleet problem nobody writes about

A reliability program is supposed to answer one question continuously: is this fleet's performance drifting in a direction that predicts failure? On a fleet of three or four aircraft, the statistical machinery built for larger operators doesn't have enough data points to trigger meaningfully. An alert level calculated the way a 40-aircraft fleet would calculate it will sit dormant on a 4-aircraft fleet not because nothing is happening, but because the math was never built for this scale.

That's not a hypothetical. It's the exact gap that shows up when an NCAA review looks closely at a small carrier's Reliability Programme Manual: the structure looks complete, the sections are all present, and the actual sensitivity to real operational drift is close to zero.

What actually works at small scale

A reliability program built for a small commercial fleet has to trade statistical elegance for operational judgment:

Component and system-level trend review has to be manual and frequent, not purely threshold-triggered — because thresholds calibrated for scale won't fire in time on a small fleet.

Every finding needs a named owner and a closure date, not just a category — a small technical team can't absorb a backlog of "monitoring" items the way a large one can.

The program has to talk to maintenance planning directly, not just report upward — on a small fleet, the person who owns reliability data and the person who owns the next check are often steps apart, and that gap is where drift gets missed.

Alert levels should be set from the fleet's own operating history, recalculated as real data accumulates, rather than imported wholesale from a manual template built for a different fleet size.

Why this is a Post Holder's problem, not just a QA exercise

When an NCAA finding lands on a Reliability Programme Manual, the temptation is to respond at the level of the document: rewrite the section, close the paperwork gap. The harder and more useful response is to ask whether the program was ever actually sized for the fleet it's meant to protect.

A reliability program that's honestly built for a small fleet won't look like a shrunk-down version of a major carrier's manual. It will look leaner, more manual in places, and more tightly wired into day-to-day maintenance decisions — because on a small fleet, there's no volume to hide behind. Every tail number carries more weight, and the program has to be built to notice that.

This is the standard Flyometrics applies when reviewing or standing up a reliability program: not "does this look like what a bigger airline has," but "does this actually catch drift on the fleet size sitting in front of us."