A maintenance program audit is a structured review of the entire maintenance function against best-practice criteria, strategy, planning and scheduling, PM and PdM, work orders, spare parts, data, KPIs, and skills, scored to reveal where the program is strong, where it leaks money, and what to fix first. It grades the whole system, not a single machine.
Most plants audit their equipment far more often than they audit the way they maintain it. The result is a maintenance department that has quietly drifted into firefighting: plenty of activity, little of it planned, and no shared picture of why. A maintenance audit fixes that by holding the function up against a known standard and turning a vague sense that "we're too reactive" into a scored, prioritized list. This guide covers what to audit, the benchmarks to grade against, how to run the audit, and it comes with a free printable checklist you can take to the floor.
What does a maintenance audit actually cover?
It covers the handful of domains that together determine whether maintenance is planned and improving or reactive and stuck. You are not just checking whether PMs get done, you are checking whether the whole system that generates, plans, executes, and learns from work is sound. The core domains:
- Strategy and criticality. Is there a maintenance strategy tied to asset criticality, or does every machine get the same generic attention?
- Planning and scheduling. Is work planned and scheduled before it is done, or dispatched as it breaks? This is the single biggest lever, covered in maintenance planning and scheduling.
- Preventive and predictive programs. Do PMs exist, get completed, and actually prevent failures? Is any predictive maintenance or condition-based maintenance in place?
- Work order process. Does every job flow through a work order with a reason, parts, labor, and a close-out, or does work happen off the books?
- MRO and spare parts. Are critical spares identified and stocked, or is every breakdown also a parts hunt?
- CMMS and data. Is there a CMMS holding clean, used data, or a system nobody trusts?
- KPIs and review. Are maintenance KPIs measured and acted on, or is nobody keeping score?
- Skills and safety. Are technicians trained for the work, and is maintenance safety, lockout/tagout, permits, built in?
Why audit the maintenance function at all?
Because a maintenance department can be busy, well-staffed, and still be losing money, and only an audit against a standard shows you where. The default failure mode is drift into reactive maintenance, which the U.S. Department of Energy's Federal Energy Management Program notes costs several times more than planned work per repair. An audit surfaces that drift as a number instead of a feeling.
It also settles arguments. "We need more people" and "we need a new CMMS" are common requests, but an audit tells you whether the real constraint is headcount, planning discipline, parts availability, or data. Spending against a scored gap analysis beats spending against the loudest voice, and it gives you a baseline to re-audit against a year later to prove the money worked.
What benchmarks should you grade against?
You grade against the commonly cited best-practice targets that separate a proactive program from a reactive one. Treat these as directional industry rules of thumb, not certified thresholds, the point is the gap between your numbers and these, not a pass/fail line:
| Metric | Reactive / typical | Best-practice target |
|---|---|---|
| Planned & scheduled work | Under ~50% | 85–90%+ |
| Reactive / emergency work | 50%+ | Under ~10–15% |
| Schedule compliance | Ad hoc | Over ~90% |
| PM completion (on time) | Often below 70% | Over ~90% |
| Wrench time (tool time) | ~25–35% | ~55–60% |
| Ready backlog | Unknown or huge | ~2–4 crew-weeks |
The single most telling ratio is planned versus reactive work. A department running over half its hours reactively cannot plan, cannot stock parts ahead, and cannot improve, because it never knows what tomorrow holds. Moving that ratio is what nearly every other improvement is ultimately for, and it ties directly to your maintenance backlog and schedule discipline.
How do you run a maintenance audit?
You run it as a short, evidence-based project: define the scope, pull real data, score each domain against the benchmarks, and turn the gaps into a ranked action plan. The discipline is in scoring against evidence, not opinion. Work through it in order:
- Define scope and criteria. Decide which lines or areas are in scope and write down the criteria for each domain up front, so scoring is consistent. The printable checklist below gives you a ready-made set.
- Pull the real numbers. Before you interview anyone, extract the data: planned-vs-reactive ratio, PM completion, schedule compliance, backlog, and downtime. Numbers beat impressions, and gaps between what people say and what the data shows are themselves findings.
- Walk the floor and interview. Verify on the floor what the system claims. Are PMs meaningful or rubber-stamped? Are critical spares actually on the shelf? Do technicians trust the CMMS? Talk to planners, techs, and operators.
- Score each domain. Rate every domain against your criteria, a simple 1-to-5 or red/yellow/green scale works. Score on evidence, and note the specific gap behind each low score, not just the number.
- Build a ranked action plan. Sort the gaps by impact and effort. Fixing planning and scheduling usually outranks everything because it unlocks the rest. Assign owners and dates, and set a re-audit date to measure the change.
Keep the whole cycle tight, a focused audit of one plant is a matter of days, not months. The output that matters is not the score; it is the short list of changes, in priority order, with owners. An audit that ends in a binder nobody opens has failed.
What are the most common audit findings?
The same handful of gaps show up in plant after plant, and knowing them tells you where to look first. An experienced auditor could almost predict the low scores before walking the floor:
- No criticality ranking. Every asset gets the same generic PM, so effort is spread evenly instead of aimed at the machines that actually stop production.
- PMs that are rubber-stamped. The completion number looks great, but the tasks are vague ("check pump") and get signed off without real inspection, so they prevent nothing.
- Schedule compliance never measured. Work gets done, but nobody tracks whether it was the planned work, so the week is really a series of interruptions with a schedule on paper only.
- Work happening off the books. Techs fix things without a work order, so the history is incomplete and recurring failures never surface in the data.
- An inaccurate storeroom. The system says a critical spare is in stock; the shelf disagrees, and the discovery happens mid-breakdown.
- A distrusted CMMS. The tool exists but the data is stale, so people work around it, which makes the data staler, a self-reinforcing loop.
Notice how these compound: no criticality ranking makes PMs generic, generic PMs get rubber-stamped, off-the-books work hides the failures that would prove the PMs are weak, and distrusted data hides all of it. Breaking the loop almost always starts with planning and clean data, which is why those two domains carry the most weight in a scored audit.
The numbers worth knowing
The case for auditing, and for climbing the maturity ladder it measures, rests on well-documented maintenance economics:
- The U.S. Department of Energy's Federal Energy Management Program O&M Best Practices guidance reports that reactive (run-to-failure) maintenance typically costs several times more per repair than planned maintenance, and that layering in predictive methods saves roughly 8–12% over preventive maintenance alone.
- Commonly cited reliability benchmarks put best-practice programs at 85–90%+ planned work schedule compliance above ~90% and wrench time near 55–60% versus a typical 25–35%, the exact figures vary by source, so use them as targets to close a gap against, not certified limits.
- Because planning is the biggest lever, the DOE FEMP guidance and most reliability frameworks agree that the highest-return first move for a reactive plant is building genuine planning and scheduling before buying new technology.
Where does the audit lead?
A maintenance audit is the starting line for a reliability program, not the finish. It gives you a scored baseline and a ranked list; the value comes from working that list, then re-auditing to prove the program moved. Run annually, it becomes the governance layer over your whole maintenance effort, the yearly checkpoint that keeps the program from drifting back into firefighting, and the natural companion to a total productive maintenance or equipment reliability initiative.
The plants that get the most from auditing make the underlying data effortless to pull, because an audit is only as good as the numbers behind it. When work orders, PM completions, downtime, and parts usage live in one searchable system instead of on clipboards and in spreadsheets, the audit stops being a data-gathering ordeal and becomes a quick read of a live picture, the same shift from scattered paper to searchable knowledge that the team in our CLS case study made. Grab the checklist below to run your first audit, and tie the findings to real floor data so next year's audit measures progress instead of re-litigating the same debates (see how Harmony keeps that data searchable).