Maintenance training hours: turning good intentions into a number

Ask a maintenance manager how much training the team gets and the answer comes back confident: "We send people on courses." Ask how many hours per person last year, and the room goes quiet.

Good intentions are not the same as a programme. Until someone actually counts the hours, there is no way to know whether the investment is steady, growing, or quietly fading as other priorities crowd it off the calendar.

What it actually measures

Maintenance training hours is the total annual hours of formal training divided by the number of maintenance employees, read as hours per person per year. You can also express it as a share of total hours worked. Formal here means classroom or hands-on instruction with a check that people learned something, covering safety, technical skills, planning, reliability techniques and the like; relevant conferences and seminars count too.

The reason to track hours rather than dollars is simple: hours are honest. A dollar of training means something different in every country and every year, but an hour is an hour everywhere, which makes this the one training measure you can fairly compare across sites, regions and time.

How to work it out

Training hours per employee = Total training hours / Number of maintenance employees

A department of 20 people completes 1,420 hours of training across the year, against 38,400 total hours worked.

Training hours per employee = 1,420 / 20 = 71 hours per year Training hours = 1,420 / 38,400 × 100 = 3.7% of total hours

At 71 hours a person, this team is within reach of the best-in-class mark, and the percentage shows how much of its working time it chose to point at getting better.

What good looks like

The figure most often quoted for best-in-class is 80 hours per person per year, though a sensible reading is 40 to 80 hours once a team is already performing well. It is worth knowing how demanding even the lower end is: in many regions only a small minority of plants manage more than 40 hours a person, so the best-in-class mark is genuinely aspirational rather than routine. A team well behind has further to climb than the steady-state number suggests; real catch-up programmes have run at around 160 hours a person a year for a couple of years to close a deep skills gap before easing back. If your team sits low, close the gap before worrying about holding steady, and set separate targets by trade, since a mechanic, an electrician, a planner and a supervisor rarely need the same diet.

Hours are an input, not the result

This is the trap that hollows the metric out. It is perfectly possible to log a healthy hour-count and teach almost nothing, the pattern some call sheep-dipping: herd a group into a room, lecture at them, and let them leave hoping something stuck. Hours measure attendance, not learning, and real learning does not happen until people practise. The plants that take this seriously test for proficiency at the end and treat training as a process rather than an event, so the number on the report reflects competence gained, not just seats filled.

Two more distortions are worth naming. The first is who gets the hours: in many organisations managers and supervisors are trained at roughly twice the rate of the people actually turning the wrenches, which is precisely backwards if the aim is reliable equipment. The second is content: generic, canned courses can pad an hour-count handsomely while doing little for craft skill. The cure for both is to build the programme from a proper skills assessment and job-task analysis, so the hours land on the right people learning the right things, and then to reassess and prove it worked.

Where it can mislead you

• Keep mandatory regulatory and annual safety training in its own bucket. It has to happen regardless, so counting it here inflates the figure and disguises how much genuine skill development is going on.
• Count only formal, documented, tested training. On-the-job coaching is valuable, but it is not the same thing and does not belong in this number.
• A healthy average can hide a neglected trade, or a workforce where the training is going to the people who need it least. Break the figure down by job type rather than trusting the single site average.
• Remember the number is a means, not an end. Hours without a check on proficiency are a comforting fiction.

Counting training hours per person is how a vague commitment to "developing the team" becomes something real: a number you can set a target against, watch over time, and answer for. What gets counted gets protected, and training is too easy to lose to leave uncounted, just make sure you are counting learning, not merely time in the room.