How to Measure Your Maintenance Backlog (in Crew-Weeks)

Maintenance backlog is all the identified, approved work that has not been completed yet. The clearest way to measure it is in crew-weeks: Backlog (crew-weeks) = Total Estimated Labor Hours ÷ Available Labor Hours Per Week. Counting work orders tells you little. Counting the labor hours behind them tells you whether you are keeping up.

Why hours, not work-order counts

A backlog of 200 work orders sounds alarming until you learn that 180 are ten-minute inspections. A backlog of 20 sounds fine until you learn that several are multi-day overhauls. Raw counts hide the size of the work, which is why mature maintenance teams measure backlog in labor hours and express it as crew-weeks, the number of weeks it would take the current crew to clear it.

The crew-weeks formula

Backlog in crew-weeks = Total Estimated Labor Hours in the backlog ÷ Available Labor Hours per week.

Available labor hours is the realistic capacity of your crew in a week, after subtracting time lost to meetings, travel, and breaks, often called the wrench-time-adjusted figure. Using gross headcount hours overstates capacity and makes the backlog look smaller than it is.

Calculate your backlog

Enter your backlog hours and crew capacity to get backlog in crew-weeks and a quick health read. The defaults match the worked example below.

A worked example

Take a maintenance crew and its current backlog: 1,200 estimated labor hours, a crew of 10 technicians, and a realistic 30 available hours per technician each week.

Metric	Formula	Calculation	Result
Weekly capacity	Technicians × hours each	10 × 30	300 h/wk
Backlog	Backlog hours ÷ weekly capacity	1,200 ÷ 300	4 crew-weeks

Four crew-weeks means that if no new work arrived, the crew would need a month to clear the backlog. That single figure is far more useful than the raw work-order count for planning staffing and spotting when you are falling behind.

The three layers: total, ready, and scheduled backlog

Backlog is not one pile. The most useful way to see it is as three layers, each a subset of the one before.

• Total backlog: every outstanding approved work order, including jobs still waiting on parts, engineering, or access.
• Ready backlog (sometimes called planned backlog): work that is fully planned, with parts on hand and ready to schedule. This is the backlog your planners actually work from.
• Scheduled backlog: ready work that has already been committed to a specific week or shift.

Tracking all three tells you more than the headline number does. A large total backlog with a thin ready backlog usually means a planning or parts bottleneck, not a labor shortage. The work exists, but the crew cannot start it.

Reading backlog by craft, age, and priority

A single crew-weeks figure hides structure that changes what you do about it. Three cuts matter most.

• By craft. Split the backlog by trade, mechanical, electrical, and instrumentation. A balanced total can hide one trade drowning while another sits idle, which is a scheduling problem, not a capacity one.
• By age. A backlog aging view shows work orders that keep getting deferred. Old jobs that never rise to the top are a sign your prioritization is not working.
• By priority. A growing pile of old, high-priority work is a different problem from a healthy buffer of routine jobs, even when the crew-weeks total looks identical.

For deciding which jobs rise to the top when you cannot do everything, see asset criticality.

What counts as a healthy backlog

A common rule of thumb is 4 to 6 crew-weeks of total backlog, with roughly 2 to 4 weeks of ready backlog, though the right range varies by site. The point is that some backlog is healthy. It gives planners a buffer to build efficient schedules instead of reacting day to day.

• Too low (under about 2 weeks) can mean you are overstaffed, or that work is not being identified and logged.
• Too high (over about 6 weeks) means work is piling up faster than you can clear it, deferred maintenance is accumulating, and risk is rising.

What makes a backlog grow

A backlog grows whenever new work is identified faster than the crew completes it. The drivers are usually some mix of the following: low wrench time, so each technician completes less in a paid hour than the headcount suggests; weak planning, so jobs sit unready while parts and permits are chased; and a high rate of corrective, breakdown-driven work crowding out planned maintenance.

That last driver is worth naming, because it is downstream of reliability. Assets that fail often generate a steady stream of corrective work orders, so a rising corrective backlog is often a reliability problem wearing a maintenance-capacity disguise. Improving the reliability of the worst offenders, tracked through MTBF, reduces the inflow rather than just clearing the outflow.

How to reduce a maintenance backlog

There are four levers, and they are not equal. The first three clear the backlog you have. The fourth keeps it from coming back.

• Clean it. Close stale, duplicate, and no-longer-needed work orders. A surprising share of an overgrown backlog is not real work, and clearing it sharpens every other number.
• Raise throughput. Improve wrench time by cutting travel, waiting, and time spent looking for parts, and plan jobs so they are fully ready before they hit the schedule. The crew completes more per week without adding people.
• Add temporary capacity for a planned catch-up. Overtime or contractors can clear a spike, but only point them at ready backlog. Throwing labor at unplanned work just creates a different bottleneck.
• Reduce the inflow. Improve reliability on the assets that generate the most corrective work, so fewer breakdowns turn into work orders in the first place. This is the durable lever. The other three are one-time.

One caution: do not aim for zero. As the healthy range shows, a backlog near zero usually signals overstaffing or uncaptured work, and it leaves planners no room to schedule well.

Where your backlog lives: pulling it from a CMMS

The raw data for backlog sits in your CMMS or EAM, systems like Maximo, SAP PM, or similar. The crew-weeks calculation needs only two things out of it: the summed estimated labor hours of all open, approved work orders, and your crew's realistic weekly capacity.

Two practical snags show up almost every time. Estimated hours are often missing or rough, which makes the numerator soft until estimating discipline improves. And weekly capacity should be the wrench-time-adjusted figure, not gross headcount hours, or the backlog will look smaller than it is.

The metrics backlog connects to

Backlog does not stand alone. It sits alongside a small set of maintenance KPIs that, read together, tell you whether your planning is working: schedule compliance, whether you completed the work you scheduled; PM compliance, whether preventive tasks are done on time; and wrench time, the share of paid time actually spent turning wrenches.

It also connects forward, into production. Deferred maintenance does not stay in the maintenance department. It eventually shows up as unplanned downtime, which is an availability loss, and availability is the first of the three factors in OEE. That is what makes backlog a leading indicator: a backlog trending the wrong way today is an effectiveness problem arriving on the floor next quarter.

Common mistakes

• Measuring by work-order count instead of hours. Counts ignore the size of the jobs and mislead.
• Including unestimated or invalid work orders. Stale, duplicate, or no-longer-needed orders inflate the backlog. Clean it regularly.
• Ignoring backlog age and priority. The total can look fine while critical work quietly ages.
• Not separating ready from waiting work. Planners need the ready backlog. Lumping in jobs stuck on parts hides what can actually be scheduled.

From measuring backlog to clearing the right work first

Backlog in crew-weeks tells you how much work is outstanding and whether you are keeping up. What it does not tell you is which jobs to do first when you cannot do them all, which deferrals are safe, and which are quietly raising the risk of a failure. That judgment depends on asset criticality, downtime cost, and failure history spread across several systems.

This is the part SteelTree handles. It connects to the CMMS and the systems around it, and where the data is thin or scattered, it captures what it needs directly from the work itself. It turns the backlog into a prioritized plan: which work protects the most production and safety, which can wait, and what to schedule this week, with the reasoning attached. And because it captures the reasoning behind each decision, the system compounds. It gets sharper at your plant the longer you run it.

See how SteelTree can transform your operational processes →