How One HVAC Company Stopped Losing Margin on Every Job That Ran Long

Every HVAC owner has felt this one. A tech gets dispatched on a job that's quoted at three hours. The invoice gets written, the customer signs off, everyone moves on. Then payroll closes for the week and somehow that same job took five hours to finish. Nobody flagged it in the moment. Nobody felt it happen. The only place it shows up is in the number that matters most: the margin on the job.

That was the exact situation for a growing HVAC company running service and install crews, the kind of shop doing steady residential and light commercial volume with a handful of trucks on the road every day. The owner wasn't short on good techs. He was short on a way to see, in real time, when a job was quietly eating into the margin he'd already priced in.

The problem with quoted time and actual time

Here's what makes this so common in HVAC and other field service businesses: the quote gets built around an estimated labor time, but nothing in the day to day workflow ties that estimate back to what actually happened on site. A tech who runs into a stuck valve, a bad access point, or a customer who wants to talk through three extra questions isn't doing anything wrong. But none of that gets captured as a signal until someone in the office reconciles hours against invoices, usually weeks later, usually as a spreadsheet exercise nobody enjoys.

By the time that reconciliation happens, the job is done, the invoice is paid, and the only thing left to do is shrug and hope next week goes better. There's no mechanism for a tech to notice mid job that they're behind pace, and there's no reward built in for finishing clean and on time. Pay is pay, whether the job took three hours or five. So there's genuinely no financial signal telling anyone, tech or dispatcher, that time is money on this specific ticket.

The owner put it plainly on a call with ShareWillow: jobs were quietly running long, and nobody felt it until the numbers came in at the end of the pay period, by which point it was too late to do anything but note it and move on. That's not a training problem. That's a visibility and incentive problem, and it's one that shows up in nearly every shop running on flat hourly pay or a flat commission rate with no tie back to actual job economics.

Why this quietly costs more than it looks like

A single job running two hours over doesn't feel like a crisis. But run that across a fleet of trucks, every week, all year, and it adds up to a meaningful chunk of margin that never shows up as a line item anywhere. It's invisible by design, because nothing in a typical pay structure is built to catch it. Most HVAC business owners can tell you their average ticket size and their close rate. Far fewer can tell you, with any confidence, what percentage of their labor margin gets eaten by jobs that ran long for no billable reason.

The job was quoted at three hours. It kept taking five, and nobody upstream ever saw it happen until payroll.

That gap between quoted time and actual time is exactly where a well built incentive plan earns its keep. Not by punishing techs for jobs that legitimately run long due to complexity, and not by pushing crews to rush and cut corners, but by giving everyone, techs and management both, a shared number to watch that reflects reality instead of a guess made weeks after the fact.

Most shops are flying blind on this without knowing it

Ask a typical HVAC owner to pull up job level margin for the past month and most can't do it without a lot of manual digging, and even then the number is only as good as whoever built the spreadsheet. Dispatch software tells you where a truck is and when a job was closed. Accounting software tells you what got invoiced and what got paid. Almost nothing in between tells you, at the moment a job wraps, whether it actually hit the margin the business needs to stay healthy.

That gap gets wider as a fleet grows. A two truck shop can sometimes catch a problem through sheer proximity, the owner is in the field enough to notice a pattern. A shop running six, eight, or a dozen trucks loses that proximity fast. Jobs happen across town, across the week, across techs the owner might not see in person for days. Without a system built to surface the margin question automatically, it simply doesn't get asked until the numbers force the conversation.

Building a plan around the number that actually matters

Instead of trying to police hours after the fact, the fix here was to build the incentive plan around a labor margin target: 20 to 25 percent, tracked at the job level instead of the payroll level. Every hour a tech books against a job rolls up against that target automatically. There's no separate spreadsheet, no manual tracking sheet taped to a clipboard, and no waiting for the office to run a report at the end of the month.

The mechanics matter here, because a plan like this only works if it's actually protecting margin rather than just rewarding speed. So the plan includes built in time and quality thresholds: a tech doesn't get credit for finishing fast if the job comes back as a callback, and they don't get penalized for a job that legitimately requires extra time due to complexity that was outside their control. The target rewards jobs that hit the margin band the business actually needs to stay healthy, not jobs that were rushed or jobs that ran long through no fault of the tech.

This is a meaningfully different approach than the two most common shortcuts owners reach for. The first shortcut is a flat hourly wage with no incentive layered on top at all, which leaves margin protection entirely up to hope and supervision. The second is a flat commission on the invoice total, which rewards closing the sale but says nothing about whether the job was actually profitable to deliver. Neither one gives a tech a reason to care about the three hours versus five hours question, because neither one measures it.

Tying pay to a labor margin target built directly into a performance pay plan changes that. The tech sees the same number leadership sees. If a job is tracking toward the margin band, that's visible in near real time instead of after the fact. If it's not, that's visible too, early enough to actually do something about it on the next job instead of just noting it in a monthly review nobody reads closely.

A few things had to be true for a plan like this to actually work in the field, and they're worth naming because they apply to almost any shop trying something similar:

• The target has to be visible to the tech doing the work, not just to the office. A margin band nobody sees can't change anyone's behavior.
• It has to reward outcomes, not just speed. A fast job that comes back as a callback should never score better than a slightly slower job that holds up.
• It has to run on data the business already has. If it requires new hardware, a new app the team has to learn, or a change in how jobs get logged, adoption stalls before it ever gets a fair test.

Why this works better than adding more oversight

It would be reasonable to assume the fix here is more supervision: a dispatcher checking in more often, a manager reviewing job times weekly, more meetings about efficiency. But more oversight is expensive, it's resented by good techs who don't need to be micromanaged, and it still only catches problems after they've already happened. A margin target built into the pay plan does the opposite. It puts the signal directly where the work happens, in real time, tied to money the tech actually sees land in their own paycheck.

That's the real shift here. Instead of a business owner trying to manually catch every job that ran long, the incentive structure itself does the catching, because it's the thing being measured and paid on every single week.

The result: margin that's visible instead of guessed at

With the plan live, the business now has a labor margin target of 20 to 25 percent tracked on every job, instead of a number that only gets estimated in hindsight. That protection is built into the ticket itself, guarded by the time and quality thresholds so it rewards genuinely efficient, quality work rather than rushed work. The owner no longer has to wonder, weeks after the fact, whether last month's margin held up. He can see it as jobs close.

That's the piece that tends to surprise owners once they see it running: the plan wasn't complicated to stand up, and it didn't require new hardware or retraining a whole crew on some new process. It required connecting the pay plan to the data that was already sitting inside the business, quoted time, actual time, and margin, and giving techs a reason to care about all three at once.

What this means if you're running a similar shop

If you're an HVAC owner, a facility manager overseeing a maintenance team, or anyone running a crew where labor hours are the single biggest variable between a profitable job and a break even one, the lesson here isn't "track more." It's track the right thing, and connect it directly to pay, so the incentive and the visibility live in the same place. A margin target only works if the person doing the job can see it and feels it land in their check. Otherwise it's just another report nobody reads.

This kind of plan tends to work well anywhere the gap between quoted and actual time quietly erodes profit: HVAC, plumbing, electrical, and plenty of other trades covered under field service and home services work. The specifics of the margin band will differ shop to shop, but the underlying fix is the same. Stop measuring after the fact. Build the target into the plan, and let the number do the work a manager used to have to do by hand.

It's also worth saying what this isn't. It isn't a call to slash pay or make techs anxious about every minute on a job. Good techs already want to do quality work efficiently, they just rarely have a clean way to see whether they're succeeding at that in real dollar terms. Giving them that visibility, tied to pay they actually receive, tends to be received as fair rather than punitive, especially when the thresholds are built to protect against penalizing anyone for a legitimately hard job.

If you're curious what a plan like this could look like for your own crew, ShareWillow builds these kinds of performance pay plans around the numbers your business already tracks, with no new hardware required to get started.