Ask a maintenance manager what percentage of their team's shift is spent actually doing maintenance and most will say somewhere between 50 and 60 percent. The industry data is consistent and significantly lower: the average is 25 to 35 percent. On a 10-person crew running at 30 percent wrench time, you have the productive output of three technicians. The other seven are on the clock — travelling between jobs, waiting for parts, chasing permits, filling in paperwork, or standing around waiting for access. None of that is the technicians' fault. Almost all of it is a planning and systems failure.
Identify the Six Time Thieves Killing Your Wrench Time — Before You Consider Hiring Another Technician
OxMaint eliminates the most common wrench time killers: pre-staged parts, geographic job clustering, mobile work orders with instructions attached, and automatic permit linking — all from one maintenance platform. Sign up free or book a demo to see the wrench time workflow live.
The Numbers That Should Make Every Maintenance Manager Stop
Wrench time — also called tool time or spanner time — measures the percentage of a technician's available shift hours spent directly performing hands-on maintenance work. Turning wrenches, replacing components, executing inspections, completing repairs. Everything else does not count.
Industry Average
25–35%
Most plants. DuPont studies and subsequent industry research consistently land here. A 10-hour shift produces 2.5–3.5 hours of actual maintenance.
Achievable with Planning
45–50%
With structured maintenance planning and scheduling in place. Equivalent to increasing a 10-person team to the output of 13–14 technicians at zero additional cost.
World-Class
55–65%
Top-performing operations with CMMS-enabled dispatch, pre-staged parts, and geographic job routing. A 57% productivity increase over the industry average.
The maths is stark. A team of 30 technicians running at 35 percent wrench time has the productive output of roughly 10.5 technicians. Move that same team to 55 percent and you recover the equivalent of 6 additional technicians — without a single new hire, without overtime, without any change to headcount. Sign up free on OxMaint to start tracking where your team's time actually goes.
Where the Other 65–75% Goes: The Six Time Thieves
Non-wrench time is not random — it clusters around six predictable failure points in the maintenance workflow. Most of them are invisible to management because nobody measures them. A wrench time study makes them visible. A CMMS makes them fixable.
01
Travel and Job Routing
Typical share of non-wrench time: 20–25%
Unoptimised job routing sends technicians across a site in random arrival order. A technician covering six jobs in a day may walk the same route four times if jobs are dispatched in the order they were raised rather than their location. On large process or manufacturing sites, this alone accounts for an hour or more of dead time per shift.
Fix: Geographic job clustering in dispatch. Group same-area jobs into one block before assigning them. A CMMS with asset location data can do this automatically on every shift.
02
Waiting for Parts
Typical share of non-wrench time: 15–20%
A technician arrives at a job, identifies the required part, discovers it is not in stock, and waits — sometimes for hours, sometimes for days. The asset sits open, the technician is unproductive, and emergency procurement adds cost on top of the delay. This is the most expensive wrench time thief on sites without pre-staging discipline.
Fix: Parts pre-staging tied to scheduled work orders. Generate pick lists before job start so parts arrive at the job, not after it begins. Planned work should never start without confirmed parts availability.
03
Waiting for Permits and Access
Typical share of non-wrench time: 10–15%
Technicians arrive at a job before the permit-to-work has been approved, or before operations has isolated the equipment. The job cannot start. The technician either waits on-site or returns to the workshop — either way, the travel time and wait time are both non-wrench. On sites with heavy PTW requirements, this single failure point can consume 45 minutes per planned job.
Fix: Permit pre-approval integrated with work order scheduling. Permits should be raised and approved before the technician's job is dispatched for that shift — not raised when they arrive at the asset.
04
Locating Tools and Information
Typical share of non-wrench time: 10–12%
Technicians leave a job mid-task to collect a specialist tool, return to the workshop for a torque specification, or visit the office to find a previous work order for context. Every return journey is dead time. Work instructions that are too generic — "inspect bearing" rather than "check bearing temperature against spec 70°C max, torque housing bolts to 45 Nm" — produce this pattern consistently.
Fix: Work orders with attached instructions, torque specs, diagrams, and previous job history — accessible on a mobile device at the asset. Technicians should have everything they need before they leave the workshop.
05
Administrative Tasks at the Asset
Typical share of non-wrench time: 8–10%
Work order completion paperwork, hand-written maintenance logs, and manual time recording — completed at the asset rather than captured digitally — consume time that is literally being spent standing still. On paper-based maintenance systems, a technician completing six jobs a day may spend 40–60 minutes on paperwork that a mobile CMMS captures in under five minutes total.
Fix: Mobile work order closure at the asset. Dropdowns for failure codes, checkboxes for inspection items, photo attachment for findings — completed in under two minutes per job, at the job, before moving on.
06
Unplanned Interruptions and Re-dispatching
Typical share of non-wrench time: 8–12%
A planned job is abandoned mid-task because a higher-priority reactive callout comes in. The technician travels to the reactive job, and the original planned job is restarted by a different technician later — re-reading the work order, relocating the parts, and re-checking the asset state. Every re-start costs 15–30 minutes of non-wrench time that never appears in any report.
Fix: Reactive-to-planned ratio management. Keeping emergency work below 15–20% of total maintenance hours minimises the frequency of planned job interruptions — which is itself one of the strongest arguments for a higher PM compliance rate.
How to Run a Wrench Time Study: Three Methods Compared
Before you can improve wrench time you need a baseline. Self-reporting and guesswork reliably inflate the number — most managers assume 50–60% before they measure. There are three practical methods for getting an accurate baseline. Book a demo to see how OxMaint's work order data gives you a continuous proxy for wrench time without running a formal study.
Most Common
Day-In-the-Life (DILO)
An observer follows a selected technician through a full shift and records activity in real time against 8–10 predefined categories: hands-on maintenance, travel, parts waiting, permit waiting, admin, training, breaks, unplanned interruptions.
Provides rich qualitative detail. Identifies specific bottlenecks within the shift structure.
Hawthorne Effect: technicians perform better when observed. One day cannot capture demand variability, emergency interruptions, and parts delays that define a normal week. Results typically inflate by 10–15 percentage points.
Most Reliable
Statistical Work Sampling
Observers take random instantaneous observations at 10–15 minute intervals across multiple shifts over two to four weeks. At each observation point, the technician's activity is recorded in one of the predefined categories. A minimum of 384 observations per technician is required for 95% statistical confidence.
Multiple observers dilute the Hawthorne Effect. Captures the full range of shift conditions including reactive callouts, parts delays, and shift handover periods that a single DILO misses.
Requires trained observers and structured scheduling. Not practical for very small maintenance teams. Does not capture technicians who are travelling or in areas the observer cannot access.
Use With Care
Self-Reporting
Technicians log their own time allocation across the same activity categories at shift end. Simplest to administer and requires no observers. Data is available immediately without study resource.
Zero observation resource required. Can cover all technicians simultaneously. Useful for broad directional assessment when formal study is not possible.
Reliably inflates results — not from dishonesty, but because technicians report defensively when they know tool time will be used to evaluate them. Consistently overstates wrench time by 10–20 points. Should never be used as the sole baseline.
Whichever method you use, introduce the study to your team correctly. The purpose of a wrench time study is not to evaluate how hard the technicians work — they are almost always working hard. It is to identify the planning and systems failures that are preventing them from doing more maintenance with the same effort. Frame it that way from the outset and data quality improves significantly.
The Planner-Scheduler Role: Why One Decision Changes Everything
The single most impactful structural change available to most maintenance managers is one that initially looks like adding cost: dedicating a resource to maintenance planning and scheduling rather than assigning it to technicians or supervisors as a secondary duty.
Without a Dedicated Planner
10 technicians
35% wrench time
= 140 man-hours productive output per 40-hour week
Planning done ad hoc by supervisors between reactive jobs. Parts not pre-staged. Job routing unoptimised. Permits raised on arrival.
→
With a Dedicated Planner
1 planner + 9 technicians
55% wrench time
= 198 man-hours productive output per 40-hour week
Net gain: 58 additional productive man-hours per week — equivalent to 1.4 full-time technicians — by converting one technician to a planning role. This is the Doc Palmer planning model applied to a 10-person crew.
The planner's job is to make sure every job starts with: the right parts staged, a complete work instruction, the permit pre-approved, and the job sequenced into a geographic cluster. None of those four things require a technician. All four of them directly attack the six time thieves above.
What Moves the Needle: The Levers Ranked by Impact
1
Parts pre-staging before job dispatch
Typical wrench time improvement: 5–10 percentage points
The single highest-impact change on most sites. When planned work orders automatically generate pick lists and parts are confirmed at the kitting stage before dispatch, parts-waiting disappears as a wrench time thief for all planned work. This alone is worth more than any other individual intervention.
2
Geographic job clustering in daily dispatch
Typical wrench time improvement: 4–8 percentage points
Grouping jobs by location before assigning them to technicians eliminates repeat travel across the site. On a large facility, this can reduce daily travel time by up to 40 percent per technician. It requires asset location data in the CMMS — something most systems capture but few use for dispatch optimisation.
3
Mobile work orders with attached instructions and history
Typical wrench time improvement: 3–6 percentage points
Eliminating office and workshop trips for information — work instructions, torque specs, previous job records, asset drawings — by attaching all relevant documentation to the mobile work order removes the most common mid-task interruption. The technician has everything they need before they leave the workshop, not after they have already started.
4
Permit pre-approval before job dispatch
Typical wrench time improvement: 3–5 percentage points
On sites with significant PTW requirements, permits should be raised and approved the day before the job is dispatched, not when the technician arrives at the asset. Integrating permit status into the work order dispatch means no job leaves the planning board without confirmed access. Jobs that cannot be permitted are deferred — not assigned and then stalled on-site.
5
Reducing reactive work ratio below 20%
Structural wrench time protection for planned work
Every unplanned reactive interruption that pulls a technician off a planned job creates two wrench time losses: the travel to the reactive job, and the re-start cost on the original planned job. Keeping emergency work below 20 percent of total maintenance hours protects planned work from interruption — which is itself the precondition for every other lever on this list to work consistently.
Stop Hiring to Cover a Planning Problem — Fix the Wrench Time First
OxMaint pre-stages parts against work orders, clusters jobs by location in daily dispatch, attaches instructions to mobile work orders, and tracks your reactive-to-planned ratio — the five levers that move wrench time without adding headcount. Sign up free or book a demo to see wrench time diagnostics from your existing work order data.
Frequently Asked Questions
What is wrench time in maintenance and why does it matter?
Wrench time — also called tool time or spanner time — is the percentage of a maintenance technician's available shift hours spent directly performing hands-on maintenance work. It excludes travel, waiting for parts, permit waiting, administrative tasks, and unplanned interruptions. The industry average is 25–35%, meaning the typical technician spends only 2.5–3.5 hours of a 10-hour shift actually doing maintenance. It matters because it is the most direct measure of whether your maintenance labour investment is being converted into productive work — and because improving it from 35% to 55% increases productive output by approximately 57% with no additional headcount.
What is a good wrench time percentage for a UK manufacturing plant?
Industry benchmarks consistently place the average at 25–35%. Top-performing operations with structured maintenance planning and CMMS-enabled dispatch achieve 55–65%. A realistic improvement target for a plant that has not previously focused on wrench time is 45–50% within 6–12 months through planning and scheduling improvements, parts pre-staging, and geographic job routing — without any capital investment in additional staff. Wrench time above 65% is possible but requires very mature planning, excellent parts availability, and minimal reactive work.
How do you run a wrench time study?
The most reliable method is statistical work sampling: trained observers take random instantaneous observations at 10–15 minute intervals across multiple shifts over two to four weeks, recording which of 8–10 predefined activity categories the technician is performing at each observation point. A minimum of 384 observations per technician is required for 95% confidence. The Day-in-the-Life (DILO) method — one observer following one technician for one shift — is the most common approach but is susceptible to the Hawthorne Effect, which typically inflates results by 10–15 percentage points above reality. Self-reporting consistently overstates wrench time and should not be used as the sole baseline method.
What is the biggest single driver of low wrench time?
Parts availability — or more precisely, the absence of parts pre-staging for planned work — is consistently the highest-impact single driver of low wrench time on most manufacturing sites. When technicians arrive at planned jobs only to discover the required parts are not in stock, the resulting wait time (sometimes hours, sometimes days on specialist items) accounts for 15–20% of total shift time across a maintenance team. Pre-staging parts from a pick list generated when the work order is planned — before the technician is dispatched — eliminates this category almost entirely for planned maintenance, which is the largest single wrench time improvement available on most sites.
Does measuring wrench time mean monitoring technicians?
This framing is the most common reason wrench time studies produce poor data and workforce resistance. The purpose of a wrench time study is to identify planning and systems failures — not to evaluate technician effort. The consistent finding across all studies is that technicians work hard and are kept from doing more maintenance by systems problems: unstaged parts, unoptimised routing, permit delays, and inadequate work instructions. Introducing a study as a systems diagnostic rather than a performance evaluation produces better data, more technician cooperation, and a far higher chance of acting on what the study finds. The improvements that follow a well-framed wrench time study consistently make technicians' jobs less frustrating — which is the point.
