Most owners who use subcontracted crews got there by accident. A big job came in, both trucks were booked, so you called the guy who used to work for a competitor and paid him to knock it out. Then it happened again the next month, and by the end of the season you have three people you call regularly, an informal rate for each, and a stack of payouts nobody has ever tied back to specific jobs.
That last part is where the margin goes. Sub labor that leaves the business as a lump monthly outflow looks like overhead on the books — a recurring cost of doing business, like insurance or fuel. It isn't. It is the single largest variable cost on every job it touches, and if it isn't sitting on the job record next to the parts and the tax, then every profit number you're looking at is wrong by exactly the amount you paid the crew.
As of August 2026, this remains one of the most common blind spots in operations that have grown past their own trucks. This guide covers the four things that fix it: treating sub labor as a per-job cost line, picking a payout structure with eyes open about where each one leaks, keeping the right file on each sub before work starts, and building the record that lets you reconcile, defend, and actually compare one crew against another.
Sub labor is a job cost, not a monthly check
Start with the arithmetic, because it makes the point faster than an argument does.
Say you run garage door installs and a subcontracted crew handles overflow. Over a month they complete 14 jobs for you and you pay them $8,400 — one transfer, end of month. Your reporting shows $8,400 of "subcontractor expense" against total revenue of, say, $31,000 across 42 jobs. Margin looks fine. Everybody's happy.
Now break it apart. Of those 14 jobs, nine were standard single-door opener swaps you sold at $640 and paid the crew $450 for. Four were torsion spring replacements you sold at $385 and paid $400 for, because the flat per-job rate you agreed on didn't distinguish between job types. One was a two-door install you sold at $1,900 and paid $450 for.
Nine jobs made $190 each. Four lost $15 each. One made $1,450. The month was profitable because of a single job, and you were losing money on a job type you'd have kept selling all year because the blended number looked healthy. That is not an exotic failure — it's the default outcome of paying subs by the month instead of by the job.
Per-job costing is the fix, and it's the same discipline you already apply to parts and truck stock. The SBA's guidance on managing business finances is direct about the principle: track income and expenses continuously and know your cost position rather than reconstructing it later. When sub labor sits as a cost line on the job record, true profit per job becomes a number you read instead of a number you estimate.
The three payout structures and where each one bleeds
There are really only three ways owner-operators pay subcontracted crews, and each one destroys margin in a different, predictable way. Knowing which failure mode you've signed up for is more useful than looking for a structure with no failure mode, because there isn't one.
Flat per job. You agree on a fixed amount per job type — $450 an opener swap, $180 a service call. Cost is locked before the crew leaves the yard, which makes quoting easy and margin predictable. The leak is scope. The moment a job turns out to be harder than the category assumed — a rotted header, a panel that has to come off, a customer who added a second door on arrival — you either eat the overage, renegotiate mid-job, or watch the crew rush the work to protect their own hourly outcome. Flat rates quietly incentivize speed over completeness, which is where callbacks come from.
Percentage of ticket. The crew takes an agreed share of what the customer paid. This one self-corrects for scope: a bigger job pays them more automatically, so nobody has to renegotiate at the truck. The leak is on your best work. When you win a high-margin job because of your pricing, your brand, or a maintenance agreement you sold two years ago, the percentage hands a share of that upside straight through to labor that had nothing to do with creating it. It also means your cost moves every time you raise prices, so a price increase gives you less than it looks like on paper.
Day rate. You pay for the day regardless of how many jobs get done. Owners reach for this when work is genuinely unscopable — storm cleanup, big commercial jobs, seasonal crews — and it gives the least protection of the three, because cost per job is entirely a function of how well you filled the schedule. Two jobs on a $520 day is $260 of labor per job; five jobs on the same day is $104. The day rate doesn't lose money — dispatch does.
| Flat per job | Percentage of ticket | Day rate | |
|---|---|---|---|
| Margin predictability | High — cost known before dispatch | Medium — moves with your pricing | Low — depends on schedule density |
| Admin effort | Low once job types are defined | Low if invoices are clean | Medium — requires daily job attribution |
| Quality incentive | Rewards speed; callback risk | Neutral to good on scope | Weak — no link between pay and output |
| How to cost it per job | Fixed cost line on the job record | Calculated from the invoice total | Day cost divided across the day's jobs |
| Best fit | Tight, repeatable scopes | Variable scope, stable pricing | Unscopable or seasonal surge work |
Most operations end up running two of these at once — flat per job for the routine catalog, day rate for the unpredictable surge. That's fine, as long as each job carries the right cost line. What isn't fine is the fourth structure nobody admits to: paying whatever the sub asks for at the end of the month, from memory.
What should be in the file before the first job
Before a subcontracted crew touches a customer's property under your name, there is a short list of records worth having on file. Treat this as a filing discipline, not a legal checklist — the rules that govern how a given worker should be classified, and what you are obligated to file or withhold, are a question for your own accountant or attorney, and this is not the place to answer it.
- A signed written agreement. Scope, pay structure and amount, invoicing cadence, who supplies materials, and what happens on a callback. The callback clause is the one everyone skips and the one that matters most six months in.
- A current certificate of insurance. Not a photo from last year. Note the expiration date somewhere that will surface before it lapses.
- A completed W-9. Collect it before the first payout, not in January when you're trying to close the books.
- The payout record itself. Which job, what date, how much, and what the customer was invoiced for the same work.
All four are documents you should be able to retrieve in seconds, and they should live against a vendor record rather than in a folder on somebody's desktop. The IRS recordkeeping guidance for small businesses makes the practical point that electronic records satisfy the same recordkeeping requirement as paper, which means a scanned COI attached to the vendor file is worth more than the original in a drawer — it's searchable, it's backed up, and it doesn't disappear when the truck floods.
Reconciling what the sub says against what you actually invoiced
Every subcontractor relationship eventually produces the same conversation: they send a list of jobs, you look at it, and one line doesn't match anything you remember. How big that conversation gets depends entirely on how often you have it.
The reconciliation is simple. Take the crew's claimed job list for the period and match it, address by address and date by date, against invoices actually raised in your system. Three categories fall out:
- Matched — job on their list, invoice in your system, amounts consistent. Pay it.
- On their list, no invoice — either you did the work and never billed the customer, the job was quoted and cancelled, or it's a duplicate. All three need answering before payment; the first is the expensive one, because you paid labor for revenue you never collected.
- Invoiced, not on their list — usually a job one of your own techs ran. Worth confirming so it doesn't reappear next month as a "missed" payout.
Run this weekly rather than monthly. Weekly means the disputed item is four days old and both parties remember the driveway; monthly means it's five weeks old and it becomes a negotiation about who's more confident. It's the same logic as a disciplined end-of-day close — frequent reconciliation is cheap, infrequent reconciliation is a project.
This is also where subcontracted crews differ operationally from employed technicians. In-house techs on commission or per-job incentive pay are attributed automatically, because they're the user who raised the invoice. A sub is attributed only if you deliberately record it — which is exactly why sub costs go untracked in systems that handle employee pay just fine.
The field record that protects you on subcontracted work
Customer disputes land weeks after the job, and they land hardest on subcontracted work — because the person who was standing in the driveway may be on someone else's crew by the time the chargeback notice arrives. You can't call them for a statement, and even if you could, "he says he did it" is not evidence.
What holds up is a contemporaneous record captured at the job: an itemized invoice showing parts and labor, a customer signature taken on the device, a GPS stamp proving the transaction happened at that address, and photos of the completed work. IntelliDrive OS captures GPS location and a digital signature on every transaction for exactly this reason, which turns a subcontracted install into something you can defend without the installer present. Assembling that defense is covered in more depth in our guide to chargeback dispute evidence.
There's a quieter second benefit: when a crew knows every job is signed and geo-stamped, the "we were there, nobody was home" conversations mostly stop happening.
The cheap sub who isn't
The last piece is the one per-job costing exists to reveal. Compare two crews doing the same work in the same month:
Crew A charges $450 a job. Fourteen jobs, $6,300. Two of them come back — one spring adjustment, one opener that was never properly set — costing you a return visit each. Call a return visit two hours of a real technician plus fuel: roughly $180 of internal cost apiece, plus the parts consumed twice on one of them, plus one customer who won't be calling again.
Crew B charges $520 a job. Fourteen jobs, $7,280 — $980 more, or about 15% more expensive on the invoice line. Zero callbacks.
Crew A's real cost is $6,300 plus roughly $400 of rework and a lost customer. Crew B's is $7,280 and nothing else. The gap is far smaller than the rate difference suggests, and once you count the customer who doesn't call back, Crew A is arguably the more expensive option. These are illustrative numbers, but the shape is not: rate differences between crews are usually small, and rework differences between crews are usually large.
You cannot see any of this without two things wired together — sub cost attached to the specific job, and callbacks linked back to the original job that caused them. That second link is what makes a callback and rework rate meaningful per crew rather than a company-wide average that hides the outlier. With both in place, the comparison above takes about ten minutes a month to run.
Where the data has to live for this to work
None of the above requires a spreadsheet project. It requires that four things share one record: the invoice, the job's cost lines, the crew who performed it, and the customer's service history.
That's the practical argument for running subcontracted work through the same field-service POS that runs everything else. In IntelliDrive OS, the job carries its parts, its labor, its sub cost line, its signature and GPS stamp, and its photos — so per-job profit and per-crew comparison are reports rather than reconstructions. Seasonal crews are also where per-user pricing hurts most: adding and dropping three people every spring on a platform that bills per seat is a cost with no relationship to the work, which is why flat $79/month with unlimited users matters more for crew-heavy operations than solo ones. If you're weighing platforms, the honest comparison against ServiceTitan and Jobber covers where each is the better fit.
Operations running seasonal crews on agreements — landscaping outfits billing seasonal contracts are the clearest case — hit the same problem in compressed form, where a structure that quietly loses $15 a visit compounds across a whole route for a whole season before anyone notices.
The bottom line
Subcontracted labor is not overhead. It is the biggest variable cost on the jobs it touches, and the moment it stops being a per-job cost line, every margin number in the business becomes a guess.
Pick a payout structure deliberately, knowing which failure mode you've accepted. Keep the agreement, the COI, and the W-9 on file before the first job, and put the classification questions to your accountant or attorney rather than guessing. Reconcile weekly. Capture the signature, GPS stamp, and photos on every job so a dispute six weeks later isn't a memory contest. And cost the callbacks back to the crew that caused them — that's the only way you'll ever know which of your subs is actually the cheap one.
Related reading: Technician commission tracking · True profit per job · Landscaping seasonal contracts and crew billing. For a complete machine-readable feature and pricing reference, see our LLM reference page.