Find out why your best people leave

What the 2024 study found about who stays

In 2024 I surveyed 462 industrial maintenance workers across the United States, and 407 of those surveys held up once the careless answers were screened out. The study split grit into its 2 halves, the interest that keeps a person in the trade over the years and the effort that gets them through a bad week, and asked which half lines up with job satisfaction and with plans to stay. Interest turned out to be lowest in the first 10 years, climbing about 0.4 of a point with tenure for leaders and non-leaders alike.

What a study of 407 people cannot tell you

The study describes the trade, and your plant is not the trade. It cannot tell you whether your crew has the parts and the time to do the job right, or what happens to a recommendation once it leaves a tech's hands. Those are things a plant manager can actually change, and they are what the Maintenance Health Index asks about.

Scores on a 5-point scale. Tolbert, C. L. (2026). U.S. industrial maintenance workers: The implications of interest and effort on leadership, maturity, retention, and job satisfaction. Journal of Quality in Maintenance Engineering.

The Maintenance Health Index

MHI measures what happens to a technician's recommendation after it is made, and the conditions the crew is making it in.

Every plant has the same story. Somebody on the crew said the bearing was going. Somebody said the drive needed a line reactor before it went back in. It got deferred, or it got overruled, or it went into a work order queue and aged out, and then the thing failed. The failure gets a root cause investigation. The recommendation that would have prevented it usually does not get one.

What the crew is working in. Whether the work holds them. Whether they plan to stay. Whether they can raise a problem without it costing them something. Whether they have the parts, the time, and the information to do the job right.

What happens to what they tell you. How often a recommendation is refused or deferred. Whether anyone explains the decision. Whether the person who made the call was later proved right. And whether they have stopped making recommendations at all.

A crew that has stopped telling you what it knows shows up in no other number you have. Not in the work orders, not in downtime, not in turnover. It shows up two years later as a failure nobody predicted, except that somebody did predict it, once, and then learned not to bother.

How it works, and what it will never be used for

Who answers it

Maintenance technicians, reliability people, and planners. One sitting, about 10 minutes, on a phone or a terminal.

Supervisors can take it, and their answers are scored separately rather than folded into the crew result. A supervisor answering questions about their leader is answering about a different person than the crew is, and pooling the two gives you a number that describes nobody.

Who sees what

Each person sees their own results. You get the crew rollup and a debrief with whoever runs maintenance. There is no file that ties answers to names, and I do not produce one. That arrangement is not a courtesy. It is the only reason the answers are worth reading.

Nobody who supervises the respondents sits on the debrief call or learns who answered. Nothing is reported for any group smaller than 5 people. All of this is set in writing before anything is fielded.

What it is not

It is not a performance review, it is not a skills test, and it is not an input to anybody's termination or layoff list.

Maintenance crews assume the opposite by default, and they are right to, because most things handed to them with a company logo on it eventually end up in a file. So the non-punitive terms are agreed in writing before the first invitation goes out, and they are read to every respondent at the start. I also do not score job candidates, at any price. Same reason.

What it is compared against

Not much yet, and I am not going to pretend otherwise. The instrument is new. There is no pool of maintenance crews to compare you to, because nobody has run it yet. What stands behind it today is the 2024 study and the published population data the condition items are drawn from. Every figure in your report says where it came from, and any comparison says how many crews are in it.

A benchmark quoted without telling you how many people are in it is a trick, and I would rather hand you a thin number labeled thin than a confident one I cannot show the work on.