Why technicians quit: what the turnover data actually shows, and what it doesn't

Here's the number that surprises most owners: construction workers quit less often than the average private-sector worker. In 2025, an average month saw 1.8% of construction workers quit, against 2.2% across all private industry¹. The trades don't have a quitting problem so much as a replacement problem: too few trained people coming in, too long to train them, and a quit that costs far more in the busy season than any percentage-of-salary rule suggests.

So the useful question about technician turnover isn't "why do techs quit more?" They don't, on average. It's "why did my good tech leave, and what did it cost?" This post works through what the government data does and doesn't show, builds the cost of one quit line by line, and maps the reasons people give for quitting to things you can actually check in your own shop.

What the turnover data actually shows

The best public source on quitting is the Bureau of Labor Statistics' Job Openings and Labor Turnover Survey (JOLTS). It doesn't break out HVAC, plumbing or electrical contractors on their own, so construction is the closest industry line. Here are the annual average monthly rates, which BLS computes as total quits over the year as a share of total employment¹²³:

Three things stand out.

Quitting cooled off. The 2021 and 2022 "great resignation" numbers have come down in construction and everywhere else. If your turnover feels worse than in 2022, it's probably not the market.

Construction separations are about layoffs. Construction loses people at a higher rate than average, but almost all of the gap is layoffs and discharges, which run close to double the private average. That's project work ending and seasonal slowdowns, and it's a fact service contractors can use: a steady year-round service job is a real selling point against new construction.

The data is national and blended. Construction in JOLTS includes homebuilders, road crews and every specialty trade together. A service shop with on-call rotations and commission pay can look very different. Use the data to reset expectations, not as your benchmark.

Tenure tells the same story

BLS's employee tenure survey shows how long people have been with their current employer. In January 2024, the median was 3.9 years for all wage and salary workers, 4.2 years in the construction industry, and 3.9 years for installation, maintenance and repair occupations. Workers aged 25 to 34, the age of most of your mid-career techs, had a median of 2.7 years⁴.

That last number is the one to plan around. A typical 30-year-old tech has been at their current job less than three years, so expect to rehire your mid-career seats regularly no matter how good you are.

The real squeeze is supply

BLS projects about 40,600 openings a year for HVAC mechanics and installers from 2025 to 2035, 42,000 for plumbers, pipefitters and steamfitters, and 72,700 for electricians, and says many of those openings will come from the need to replace workers who leave the occupation or retire⁵. Set against roughly 440,900 HVAC jobs in 2025, that's about one opening a year for every eleven jobs.

That last figure is from the Federal Reserve's survey of small employers: 46% named hiring or retaining qualified staff as an operational challenge⁶. When a seasoned tech walks, the replacement usually isn't a seasoned tech. It's a helper you have to train.

What one quit really costs: build it, don't borrow it

You'll see three numbers quoted for the cost of turnover, and they disagree by a factor of ten:

None of them is wrong. They measure different things in different kinds of jobs. For a service technician, the biggest cost is usually one none of them was built for: the gross profit the empty truck doesn't produce while you look for a replacement. That depends on the season more than the salary.

So build the number yourself. Here's a worked example with invented inputs. Replace each with yours.

Say a $30-an-hour service tech producing $4,000 a week in gross profit leaves, and it takes eight weeks to get someone new on the truck.

• Vacancy cost. In June, the rest of the crew can absorb maybe half the work. The other half gets turned away or pushed out: 8 weeks × $4,000 × 50% = $16,000.
• Overtime to cover. The crew picks up 20 extra hours a week for 8 weeks. The overtime premium alone (half of $30) is $15 an hour: $2,400, more if commission raises their regular rate.
• Recruiting. Use SHRM's cross-industry figure as a placeholder: $4,700.
• Ramp-up. The replacement costs more than they produce for weeks. The ride-along ladder example puts the first eight weeks of a new hire at roughly $7,600 to $12,800 of negative contribution. Call it $10,000.

That's about $33,100, a little over half of the $62,400 salary. Now run the same quit in November, when the crew can absorb 80% of the work and nobody needs overtime: vacancy drops to $6,400, overtime to zero, and the total to about $21,100.

The practical lesson: the cost of a quit depends more on when it happens than on what the person earned. That's why the conversations below are worth having in March and April, before your techs field their summer offers.

Why technicians say they quit

The cleanest recent data on reasons comes from Pew Research Center's 2022 survey of workers who quit a job in 2021. Majorities said low pay (63%), no opportunities for advancement (63%) and feeling disrespected at work (57%) were reasons, major or minor, they left. Also high on the list: child care problems among parents (48%), lack of flexibility in hours (45%), lack of good benefits (43%) and working too many hours (39%)¹⁰.

That's a survey of all workers, not technicians, and people can give more than one reason. But every one of those reasons has a version in a service shop, and most of them show up in data you already have.

The useful thing about this table is the third column. You can spend an afternoon on it and know which of your techs is most at risk, before they hand you two weeks' notice.

On pay, link the conversation to real numbers. Our salary guides for HVAC technicians, plumbers and electricians have the BLS percentiles by state and metro and what a tech costs once taxes and benefits are added. If a commission plan is part of the friction, how to pay technicians covers the overtime math most shops get wrong.

Pay transparency laws: your range is public now

A growing list of states require employers to put a pay range in job postings. That changes retention, not just hiring. When your competitor posts "$32 to $42 an hour" for a service tech, your techs see it. If your best tech makes $31, the posting is the start of the conversation you didn't have.

As of September 2026, these states require pay ranges in postings (thresholds are employees counted under each law; check the details for your state)¹¹:

Several of these also require a description of benefits, and some cities have their own rules. This table is a starting point, not legal advice.

Three things to do with it, whether or not your state is on the list:

• Write your own ranges down before you post. If you can't defend the gap between your highest and lowest paid tech in the same role, fix it before a posting makes it public.
• Read your competitors' postings twice a year. In covered states, they're the best local pay data you'll get, and more current than any survey.
• Tell your current techs where they sit. A tech who hears their range from you is less likely to be surprised by a competitor's ad.

Stay interviews: ask before they leave

Exit interviews tell you why someone left. Stay interviews ask the same questions while you can still fix the answer. SHRM's guidance, adapted from consultant Richard Finnegan's work, suggests a short set of questions: what people look forward to at work, what they're learning and want to learn, why they stay, when they last thought about leaving and what prompted it, and what their manager could do to make the job better¹². That's practitioner advice, not a controlled study, but it's cheap and the questions are good.

For a service shop, we'd run it like this:

1. 1

   Pick the season

   Hold stay interviews in March and September, before the busy seasons when other shops start recruiting.

2. 2

   Owner or service manager, one on one

   Not the dispatcher, and not in the truck between calls. Thirty minutes, paid time, somewhere quiet.

3. 3

   Ask five questions and mostly listen

   What do you look forward to? What do you want to learn next? Why do you stay? When did you last think about leaving, and why? What's one thing I could change that would make the job better?

4. 4

   Bring their numbers

   Their overtime share, on-call count and the mix of calls they got over the last 90 days. If they raise fairness, you have the facts in front of you.

5. 5

   Commit to one change, with a date

   Not five. One thing they asked for that you'll do, and when. Then do it before the next interview.

A retention checklist built from the data

The order matters: the first items cost little and catch most of the risk.

Software won't keep a tech who's underpaid or disrespected. It can make the pay part checkable. In TradesWork, techs see every performance pay line on their phone with the math spelled out, and a tap on Question this turns a pay question into an office task that gets answered in writing, which is the opposite of a payday argument. See performance pay for how that works.

If you're replacing techs faster than you can train them, the longer fix is growing your own through a registered apprenticeship, paired with a structured ride-along program so new hires produce sooner.