Maintenance Work Order Best Practices: What to Track

What makes a good work order

A work order can tell you too little or bury the useful details. “Fixed pump” leaves the next tech guessing; a long form full of unused fields slows everyone down. Start with these details and add what the job and your site's procedures require:

1. Clear title — Describes the problem or task in plain language. “Hydraulic leak at cylinder #3 on Press 2” beats “Press 2 issue” every time.
2. Asset — Which machine, vehicle, or piece of equipment. Not “the one in the back corner” — the actual asset name or ID.
3. Owner and priority — Name who is responsible for the next action. RunTight uses Low, Medium, High, and Critical. Agree on what those levels mean for your site, considering safety, equipment condition, and production impact; a priority label alone is not a response deadline.
4. Due date — Every work order needs a deadline. Open-ended tasks get done “eventually,” which means never.
5. Checklist or steps — What needs to happen, in order. This turns tribal knowledge into a repeatable process.
6. Time tracking — How long did the work actually take? This is the single most undervalued field on a work order.
7. Failure code — What did you find? Bearing failure, electrical fault, wear, contamination. Use consistent categories, distinguish symptoms from confirmed causes, and leave an unknown cause for investigation rather than guessing.
8. Completion record — What was repaired, what was checked, which parts were used or returned, and what still needs follow-up? Record the result as well as the activity.

The five mistakes that make work orders useless

If your team dreads filling out work orders, the problem isn't laziness — it's that the work orders are designed poorly. Here are the five most common mistakes:

1. Vague titles

“Machine down” tells you nothing six months from now. When you're trying to figure out why Press 2 has been down four times this quarter, you need specific titles that describe the actual symptom or failure. Train your team to answer: what is broken, and where?

2. No time logging

Without time data, it is harder to estimate labor cost per asset or plan the next repair. Record actual labor as accurately as practical, following your site's time-recording policy. If you have to estimate, label the estimate; an unknown duration is not the same as zero. Compare jobs with similar scope and conditions before drawing conclusions about technician performance.

Use one consistent rule: log hands-on labor as the primary duration, then note material delays separately. For example, “45 minutes troubleshooting, 30 minutes replacing the sensor, 2 hours waiting for the line to be released” is more useful than one duration that mixes labor and waiting. Record total equipment downtime separately if you need to measure the effect on production.

• Short inspection: record the time spent without rounding a real task away
• Repair or planned PM: use a timer or a recorded start and finish, and account for pauses
• Multiple technicians: record each person's labor; two people working for one hour contribute two labor-hours
• Waiting on parts or production: capture it as a delay note or downtime reason

3. Skipping failure codes

Failure codes feel like extra work when you're closing a ticket at 11 PM after a breakdown. But consistent categories make repeated problems easier to find. Start with the observed failure and a useful note. If the cause is still unknown, flag it for follow-up; a guessed code can send your next PM change in the wrong direction.

4. Too many required fields

Every field you add to a work order form is friction. If a tech has to fill out 20 fields to close a 15-minute task, check which answers someone actually uses. Keep routine entry short, while retaining the records your site requires for the task, such as readings, permits, parts, or sign-off. A short form must still capture the evidence needed to accept the work.

5. No review step

Reviewing completed work gives a supervisor a chance to check the result, identify follow-up work, and coach techs on useful notes. Set review depth according to the job: a routine inspection and a major repair need different attention. A record review also does not replace any physical checks required before equipment returns to service.

How work order data drives decisions

Review the history regularly as it builds. A busy asset may show a repeated fault quickly; an infrequently used machine needs a longer observation period. Check record completeness and operating conditions before treating a pattern as a trend.

Pareto analysis of failure codes

Pull your failure codes for the last quarter and rank them by frequency. Look for failure codes that account for a large share of your own breakdowns. Maybe it's bearing failures on your conveyors, or electrical faults on your oldest press. Once you see the pattern, you can build targeted PMs to prevent those specific failures.

This is how reactive maintenance teams become proactive ones. You're not guessing at PM schedules — you're building them from actual failure data from your own equipment.

Labor hours by asset

When you track time on every work order and apply an appropriate labor rate, you can estimate maintenance labor cost per machine. That helps with the question every plant manager eventually asks: “Should we repair or replace this machine?” For example, if Machine A consumes 40 hours of maintenance labor per month while Machine B takes 5, investigate the difference. Compare their workload, age, downtime, parts cost, and replacement options before deciding.

Repeat failures

If, for example, the same asset generates the same repair request three times in two months, look beyond the latest repair. Check whether the records describe separate failures, repeat symptoms, or duplicate requests. Reviewing work-order history helps surface these repeats so you can investigate their cause.

The work order lifecycle

Use status to describe the work's actual condition. In RunTight, the statuses are Open, In Progress, On Hold, Completed, and Skipped. Review and approval are recorded separately; they are not extra work-order statuses.

1. Open — The work order is created with a clear title, assigned to a tech, and given a priority and due date. This is the “backlog” stage.
2. In Progress — The work has started. Record the labor actually spent; the status alone does not tell you how long someone worked.
3. On Hold — Work is paused. Record what is blocking it, who owns the next action, and when it should be checked again. Waiting for parts should not look like a completed repair.
4. Completed — The work is done. The tech logs time spent, selects a failure code (for reactive work), and adds any notes about what they found. Photos are worth a thousand words here.
5. Skipped — Keep work that was not performed distinct from completed work, and explain the decision. Skipping a task does not demonstrate that the equipment is safe or the maintenance requirement has been met.

After completion, a supervisor or workspace admin can use Supervisor Review to record that the completed order was reviewed. Its status stays Completed. Check the labor, result, parts, and outstanding follow-up before marking the review. Where an approval workflow is enabled and an order requires approval, resolve that approval before starting or completing it; approval to do the work is different from review of the finished record.

Once a job has been reviewed, technicians cannot change its work record or associated parts, meter readings, and downtime entries, with one exception: technicians can still stop a running downtime event on a reviewed job that has not been archived. A completed job's details are not edited in place, so if you find an error after review, add a comment on the work order that says what was wrong and what the correct value is. Anyone in the workspace can comment, and the original entry stays in the history alongside the correction. If the mistake means work still needs doing, open a new work order for the asset rather than reusing the finished one. Supervisor Review itself is recorded by an admin or supervisor.

Record parts used and unused returns

Planned parts are a preparation list, not proof of consumption. Record what was issued and return unused items so the work order reflects what the job used. In RunTight, issuing a catalog part reduces stock on hand; recording its return adds the returned quantity back. Enter a whole return quantity no greater than the amount still issued, then check that the return saved before moving on.

A part entered by name without a catalog item can still be recorded on the work order, but it does not change catalog stock. Catalog stockroom features depend on your workspace's plan and access. Keep unknown costs distinct from confirmed zero cost, and update the record when the cost is known.

Archive obsolete work without recording a false completion

A duplicate request, retired asset, or cancelled job does not mean the repair was completed. Keep those decisions separate from completed maintenance so the history remains useful. Record why a request no longer belongs in the active queue and refer to the retained work order when combining duplicates.

In RunTight, a workspace admin can archive a work order with a reason and restore it if the decision changes. Archiving keeps the record available while removing it from active work. Before archiving, check for outstanding work that still needs an owner; tidying the list must not hide an unresolved equipment problem.

Getting your team on board

The hardest part of work order management isn't the system — it's the habit. Here's what works:

• Start with a manageable pilot — One area or type of work can help the team learn the system. Keep existing PMs, inspections, and required records running while you move the workflow over.
• Make entry practical — Watch a tech document a real job on a phone. Remove duplicated entry and confusing fields without cutting the details needed to accept the work.
• Show the value — At a regular team review, show repeated failures and labor-heavy assets from the records you have. Explain which PM or planning decision the data helped you make.
• Coach incomplete records — Help techs improve notes while the job is fresh. Resolve missing information needed for safe handback or required records before accepting the work.

The bottom line

A useful work order supports the records your site needs and builds a history you can act on. A clear symptom, honest time entry, and specific completion note help the next person understand what happened. Keep the form simple, make it mobile, review the data, and act on what it tells you.

Work order FAQ