Metrics
The 5 components of MTTR: where your team really loses time
MTTR is more than repair time. Learn the five stages that make up the response to a failure, how to measure each one and how your team can reduce it.
When MTTR goes up, the usual reaction is to put pressure on the technician. But if you open up the clock on a work order, you'll find that the repair itself is often one of the smallest parts of the downtime.
MTTR, mean time to repair, is calculated as total repair time ÷ number of corrective repairs. The math is explained in how to calculate MTBF and MTTR. The focus here is different: what sits inside that time, and how to shrink each piece.
What's inside MTTR
Between the moment equipment fails and the moment it runs again, five different stages go by, each with its own owner and its own causes:
- Detection: from the failure until someone logs a service call.
- Dispatch and travel: from the service call until the technician is at the asset.
- Diagnosis: from arrival until it's clear what needs to be done.
- Waiting for parts and logistics: from diagnosis until the part, tools and clearance are in hand.
- Repair and test: from the start of the work until the asset is released to run.
A single MTTR mixes all of this together. Two teams with the same MTTR can have opposite problems: one is slow to arrive, the other waits for parts. The fix for one doesn't help the other.
To separate the five stages, each work order needs seven timestamps: failure, call opened, dispatch, arrival, end of diagnosis, start of work and release. With them, each stage is a simple subtraction.
Example: the clock on a field work order
Take a fictional service call at an elevator maintenance company: the passenger elevator in a residential building has stopped between floors. The building manager noticed it at around 7:40 a.m.
| Timestamp | Time | Stage | Duration |
|---|---|---|---|
| Failure noticed | 7:40 a.m. | — | — |
| Service call opened | 8:12 a.m. | Detection | 32 min |
| Technician dispatched | 8:35 a.m. | Dispatch (queue) | 23 min |
| Technician on site | 9:20 a.m. | Travel | 45 min |
| Diagnosis complete | 9:55 a.m. | Diagnosis | 35 min |
| Part on site, work starts | 11:40 a.m. | Waiting for parts | 105 min |
| Elevator released after test | 12:30 p.m. | Repair and test | 50 min |
| Total downtime | 290 min |
Dispatch and travel added up to 68 minutes. Of the nearly five hours the elevator was out of service, the repair took 50 minutes. Waiting for the part, on its own, took more than twice that. Pushing the technician to work faster would change almost nothing here.
A note on definitions: some companies start counting MTTR when the service call is opened, others when the failure happens. Both are valid as long as you pick one and stick with it. Counting from the call hides detection, so record both times whenever you can.
The five stages, one by one
1. Detection
What it is: the gap between the failure happening and it becoming a logged service call. It includes the time until someone notices and the time until that person tells the right people.
How to measure it: work order opening time minus failure time. The failure time comes from the requester's report, an alarm or a monitoring system log.
How to reduce it: provide a single, easy channel for opening service calls, such as a QR code on the equipment itself. Where there are sensors, alarms can open the work order automatically. And put an end to requests that sit in someone's text messages until they get passed along.
2. Dispatch and travel
What it is: the time to decide who will respond and for that person to reach the asset. In a plant, it's a few minutes' walk. For field teams, such as HVAC, elevators and telecom, travel between sites is often the biggest block of time on the work order.
How to measure it: dispatch minus opening (queue) and arrival minus dispatch (travel). Logging arrival on site from the technician's app avoids the "from memory" time written down at the end of the day.
How to reduce it: clear priority rules, so dispatch doesn't depend on someone checking the queue. Dispatch the nearest qualified technician, not the first one available. Routes that group calls in the same area.
3. Diagnosis
What it is: the time between the technician arriving and knowing what to do. This is where individual experience counts the most.
How to measure it: end-of-diagnosis time minus arrival time. Ask the technician to mark the moment they identified the probable cause, before they start taking things apart or ordering parts.
How to reduce it: asset history available on site: what failed before, what was replaced, what fixed it. Diagnostic flows for the most common failures. Diagrams and manuals on the phone, available even with no signal, which makes all the difference in machine rooms and basements.
4. Waiting for parts and logistics
What it is: everything that stands between diagnosis and the work itself: parts, special tools, area clearance, energy lockout, quote approval, support from another team.
How to measure it: start of work minus end of diagnosis. If you can, record the reason for the wait from a fixed list (part, approval, access, tool) so you can total it by reason later.
How to reduce it: minimum stock of the parts that show up most often in corrective work, including in the field technician's vehicle. Approvals with defined spending limits. For jobs that require clearance, arrange access before dispatching.
5. Repair and test
What it is: carrying out the fix and checking that the asset is performing its function again.
How to measure it: release time minus start of work.
How to reduce it: standard procedures for recurring repairs, part kits by type of job, cross-training. And don't skip the test: releasing without testing often brings the same failure back a few days later.
Field teams vs. the plant
In manufacturing, the technician is a few yards from the asset, and the biggest stages tend to be diagnosis and waiting for parts. For teams that serve many locations, the balance shifts: travel and logistics take over, and every trip without the right part means a second visit.
That's why, for field teams, the most useful question is: how many work orders were resolved on the first visit? Every return trip counts travel twice and often hides a diagnosis made without history or a vehicle without stock.
Where to start
- Make sure every corrective work order has all seven timestamps, recorded as things happen, not at the end of the day.
- After a month, add up each of the five stages across all work orders.
- Tackle the biggest block first. Often it isn't the repair.
A CMMS like taskfield records these timestamps from the technician's app, including offline, and shows the breakdown per work order and per period, so you can see where the time really goes.
Summary
- MTTR adds up five stages: detection, dispatch and travel, diagnosis, waiting for parts, and repair and test.
- With seven timestamps per work order, each stage becomes a simple subtraction.
- In the elevator example, the repair was 50 of the 290 minutes of downtime; waiting for the part was 105.
- For field teams, travel and logistics usually weigh more than the fix itself.
- Reduce the biggest block first, which is rarely the repair work.
Calculate it now. Run the numbers with your own data in the free calculators:
