How to Hire and Keep HVAC Technicians (When Everyone Is Short-Staffed)

Most owners treat turnover as an annoyance rather than a line item, because the cost never shows up on a single invoice. It should. Replacing a technician commonly runs 20 to 30 percent of their annual pay once you total recruiting, interviewing, onboarding, and the productivity ramp — call it $14,000 to $21,000 on a $70,000 tech. That is the visible half.

The invisible half is bigger. A truck sitting idle is not just an unpaid wage, it is lost billable hours. If that tech was billing six hours a day at a $95 shop rate, every week the seat stays empty is roughly $2,850 in revenue that never gets booked — and it does not come back, because those calls went to a competitor. Three months to fill a seat is comfortably a $35,000 hole once you add the replacement cost on top.

Run that number for your own shop once. It reframes every retention decision that follows. A $3,000 raise, a $1,500 training budget, or a $200-a-month tool subscription all look very different when the alternative outcome costs five figures.

The instinct in a tight labor market is to write a job ad for a seasoned journeyman with five years of residential experience and their own hand tools. So does every other shop in town. You are now in a pure bidding war against contractors with bigger balance sheets, and you will lose it — or you will win it by overpaying for someone who leaves the moment a higher bidder shows up. A technician you bought with money can be bought back with money.

The shops that solve this permanently hire for aptitude and build the technician in-house. Mechanical curiosity, reliability, a clean driving record, and the ability to talk to a homeowner without making them nervous are the traits you cannot teach. Superheat and subcooling you can teach. Career changers out of auto repair, appliance work, the military, and even kitchen work routinely make excellent techs, and they arrive without the habits and salary expectations of someone who has been through three shops in four years.

This is a real trade-off, not a free lunch: an apprentice is a cost center for months before they are a profit center, and you need a senior tech willing to be shadowed. But you are trading money you do not have for time and structure you can actually control — and the tech you built is far less likely to leave, because they associate their career with your shop.

Most HVAC job ads are a list of demands: five years experience, EPA certification, valid license, must lift 50 pounds, must pass background check. Read that as a technician who already has a job. Nothing in it tells them why their life gets better if they take yours, and the best techs are not desperate — they are comfortable and need a reason to move.

Write the ad to the tech, not to your HR file. Lead with pay as an actual range, not "competitive compensation," which every experienced tech reads as "below market and we are hoping you will not ask." Then answer the questions they are actually weighing: How old are the trucks? Do I get one to take home? Am I on call every third week or every other week? Do you send me out with the tools and equipment to do the job right, or do I buy my own? Is there a path from install to service, or service to lead?

Be honest about the hard parts too. Summers are brutal in this trade and everybody knows it — a shop that says "July and August are long, here is exactly what overtime pays and here is how we protect your weekends the rest of the year" reads as credible. A shop that pretends the busy season is not hard reads as a shop that will surprise you.

A good technician who puts out feelers is off the market in roughly a week. Large operations cannot move that fast — they have recruiters, panel interviews, scheduling committees, and approval chains. You are the owner. You can call the candidate back in an hour, get them in the truck tomorrow, and make an offer on the spot. That is a genuine structural advantage and most small shops waste it entirely by taking four days to return a call.

Compress the process to days, not weeks: same-day callback, a short phone screen, a working interview where they ride along or diagnose a real unit in your shop, and an offer. The ride-along is the highest-signal step in the whole process — you will learn more watching someone diagnose a bad capacitor and talk to a homeowner for two hours than from any number of interviews. Pay them for that day. It costs you a few hundred dollars, it signals that you value people's time, and it filters out anyone who was never serious.

And keep the process running when you are fully staffed. Shops that only recruit when someone quits are always hiring from a position of desperation, which is how you end up with the tech you did not want at a wage you cannot support. Take the coffee meeting with the good tech from the supply house even when you have no open seat. When a seat opens, you already know who to call.

Pay has to be at market — if you are below it, nothing else in this guide will save you, and no amount of culture talk fixes an underpaid tech. But once you are at market, more money is a surprisingly weak retention tool. Techs rarely leave a shop they like for a 5 percent raise. They leave because they feel undervalued, they are set up to fail, or they cannot see where their career goes from here.

What actually retains technicians is closer to the ground than compensation. It is a truck that is not falling apart. It is the right tools and instruments, provided by the shop, so they are not fighting the job. It is a dispatcher who does not send them across town and back twice in one day. It is being told about schedule changes rather than discovering them. It is a boss who backs them when a customer is unreasonable. None of these are expensive; all of them are noticed daily.

This is why routing and dispatch quietly show up as a retention issue. A tech who spends two hours a day in windshield time is earning less on commission and getting home later, and neither of those is your fault in their mind — it is the office's fault. Tightening scheduling and dispatch typically recovers one to two billable hours per tech per day, which pays the tech more and gets them home earlier at the same time. That is a rare lever that improves margin and morale together.

Most technician turnover is decided far earlier than it happens. A tech who shows up on day one to no truck assignment, no introduction, no schedule, and a "just ride with Dave this week" has already started quietly wondering whether this was a mistake. The shops with the worst retention almost always have no onboarding at all — the new hire is thrown into the busy season and left to sink or swim.

A structured first 90 days does not need to be elaborate. Week one: ride-alongs with your best tech, not your most available one, plus truck setup, systems, and paperwork. Weeks two through four: they run simple maintenance and easy service calls solo, with a senior tech reachable by phone. Month two: harder diagnostic work, reviewed at the end of each day. Month three: a real sit-down about where they are strong, where they need work, and what the next step in their career looks like.

Two details make the difference. Assign one specific person as their mentor — "everybody is here to help" means nobody is. And check in weekly for the first month, in person, asking a real question rather than "how is it going?" Ask what surprised them, what is frustrating, and what they still do not understand. New techs will tell you exactly why they are about to quit if you ask them in week three. Almost nobody asks.

The most dangerous employee in your shop is your best technician at year four, who is excellent, fully booked, and has no idea what the next five years look like other than more of the same. That is the tech who eventually decides the only way up is to buy a van and start a competing shop — and they will take your customers with them, because those customers are loyal to them, not to you.

Give the career path a shape and put it in writing: apprentice, installer, service tech, senior tech, lead, and then the branches — field supervisor, sales, or trainer. Each rung needs stated requirements (certifications, callback rate, close rate, tenure) and a stated pay band. What kills retention is not the absence of a promotion, it is the absence of a visible, achievable path. A tech who can name exactly what they must do to earn the next $8,000 will work toward it. A tech who cannot will start browsing.

For your genuinely irreplaceable people, get creative before they force the issue: a slice of profit on the install side, a real bonus tied to margin rather than revenue, a four-day summer schedule in the shoulder season, or the training budget to specialize in commercial or controls work. Any of these costs less than losing them and then competing against them.

If you are going to build technicians rather than buy them, training stops being a perk and becomes core infrastructure. Structured training platforms with video courses, VR simulation, and skill assessments run roughly $35 to $150 per user per month, which is a rounding error against the $14,000 to $21,000 it costs to replace one tech. The catch is that these platforms only work with management commitment — blocking real, paid hours for training and reviewing progress. Buying seats nobody uses is the most common way this money gets wasted.

The second place tooling pays back is consistency in the field. Guided diagnostic tools that walk a tech through proper testing and generate quotes from real measurements — typically $30 to $120 per tech per month, plus $800 to $2,000 in wireless probes — do something training alone cannot: they put a floor under how badly a newer tech can do the job. Your year-two tech tests the system the same way your year-ten tech does, and the customer gets the same professional report either way. That is what makes hiring for aptitude survivable.

The third is coaching, and it is the one most owners underuse. The gap between a top performer and a middle performer is often two to three times in revenue per ticket on the same opportunities, and most shops have no systematic way to close it — the owner simply does not have time to ride along with everyone. AI coaching platforms that analyze call recordings, score them against a playbook, and give each tech specific feedback run about $300 to $800 a month for a shop with 5 to 25 techs, and shops report 15 to 30 percent improvements in revenue per ticket within 90 days. Used honestly, it is coaching. Used as a surveillance tool to punish people, it will cost you the techs you were trying to develop — so tell your team exactly how it works and what it is for before you turn it on.

You cannot fix what you do not measure, and most shops measure none of this. Track time-to-fill (how many days from posting to a tech in a truck), first-year retention (what percentage of hires are still with you at twelve months), and revenue per tech (are your people getting more productive, or just busier?). Review them quarterly.

If time-to-fill is long, your problem is the ad or the speed of your process. If first-year retention is bad, your problem is onboarding — the hires are fine, the landing is not. If revenue per tech is flat while headcount grows, you are hiring bodies rather than developing technicians, and no amount of recruiting will fix that. Each number points at a different fix, which is exactly why the shops that track them stop guessing.