Labor is the line item that will make or break your job. Materials get bought once and they cost what they cost. Equipment is a rental you can predict. But labor bleeds a little every single day, and if you're not counting it, you won't find out until the sub's final invoice lands three weeks after they've demobed and you have no way to prove they didn't put in the hours they're billing for.
The whole point of tracking subcontractor labor isn't to catch people cheating, though it does that too. It's so you can answer three questions honestly at any moment: Who's actually on my site? Are they keeping pace with the plan? And is the work I'm paying for showing up on the wall? If you can't answer those, you're not running the job — the job is running you.
Start with headcount, because that's where the lie lives
Every foreman will tell you they're "fully staffed." Go count the trucks in the lot at 7:15 and you'll find out what fully staffed actually means. The single most useful number you can capture every morning is a headcount by trade, and the single most useful comparison is planned versus actual.
Here's why it matters. Your three-week look-ahead assumed fifteen electricians to rough-in the third floor by Friday. Ten showed up Monday, ten showed up Tuesday. Nobody said a word. You don't find out you're behind until Thursday when the drywall crew is standing around waiting on a wall that isn't ready. If you'd logged the shortfall Monday morning, you'd have had four days to lean on the electrical super, adjust the sequence, or pull the drywall crew to a different area. Instead you ate two days of standby and a schedule slip.
The fix is boring and it works: a daily manpower log, by sub, by trade, entered before the coffee's cold. When your weekly work plan says a crew size and the field says something else, that gap is the earliest warning signal you get. Treat a chronic no-show pattern as a red flag on that sub's ability to make their milestones — because it is one.
Tie the hours to the activity, not just the day
A daily headcount tells you who was there. It doesn't tell you what they accomplished. The step that separates real labor tracking from a glorified sign-in sheet is connecting hours to a specific activity — the same activities that live on your schedule.
When a sub tells you an activity is 80% complete, that claim should be backed by labor. If the drywall hang on Level 2 was budgeted at 400 man-hours and they've burned 360 to get to a claimed 80%, the math is telling you the truth even if the sub isn't: they're going to blow the budget, and probably the date. Percent-complete claims are opinions. Hours against a known budget are facts.
This is also where your estimating gets better over time. Once you've tracked that a 5,000 SF unit of metal stud framing actually took 320 hours on the last three jobs — not the 260 the estimator assumed — you stop planning durations off a hopeful spreadsheet and start planning them off what your crews really do. That feedback loop is worth more than any single project's data.
Productivity is a rate, and rates are where the money is
Take the work put in place and divide it by the hours it took. That's productivity, and it's the number that actually predicts whether you're going to make money. Square feet of drywall per man-hour. Linear feet of pipe per man-day. Cubic yards placed per crew-hour. Pick the unit that matches the trade and track it.
What you're hunting for is variation. One framing crew doing 12 SF per man-hour and another doing 8 on the same building isn't a rounding error — that's a 50% spread, and it's telling you something about supervision, layout, material staging, or crew mix. Sometimes the slow crew is working a harder area. Sometimes the fast crew is cutting corners you'll pay for at inspection. Either way, the number gives you a reason to walk over and look, instead of finding out at punch.
A word of caution earned the hard way: don't manage productivity by beating on the crew. Half the time the crew isn't the problem. If they're waiting on layout, chasing materials from a gang box on the wrong floor, or working around another trade that's in their way, no amount of hustle fixes that. The rate drops for a reason. Your job is to find the reason, and usually it's a coordination failure upstream — which is exactly what a good look-ahead is supposed to prevent.
Watch overtime like a smoke detector
Overtime isn't automatically bad. Planned, priced overtime to hold a critical date is a legitimate tool. Unplanned, recurring overtime is a symptom, and the symptom is almost always understaffing or a broken sequence.
When the same trade is running Saturdays three weeks in a row, that's not a burst of effort — that's a crew trying to make a straight-time schedule work with too few bodies, and it's costing you a premium to do it. Flag it. The math is simple and ugly: forty hours of overtime at time-and-a-half is the labor cost of sixty straight-time hours, and tired crews in week three make mistakes that show up as rework in week four.
If you're being asked to compress a schedule, model the overtime before you commit. Know what the acceleration actually costs in premium hours and lost productivity — crews on a sixty-hour week don't produce at the same rate as crews on forty — so you can decide with your eyes open instead of discovering the bill later.
Where scheduling and labor tracking actually meet
Labor data is only worth collecting if it changes a decision, and the place it changes decisions is your short-interval plan. This is the honest connection to look-ahead scheduling, and it's a two-way street.
Going forward: your labor history tells you how many bodies an activity really needs, so the crew sizes in your weekly work plan stop being guesses. Going backward: your actual hours confirm whether the work you planned last week actually happened. An activity you slid to "in progress" should have hours behind it. If it doesn't, someone's reporting progress that isn't there.
A tool like LookAheadWall is built around that weekly, location-based plan and the trade-flow sequences that connect one crew's work to the next. The value of pairing labor data with that kind of look-ahead is that a manpower shortfall stops being an abstract worry and becomes visible against a specific activity, in a specific location, on a specific day — the exact place a superintendent can still do something about it. The schedule shows you what's supposed to happen and in what order; the labor numbers tell you whether the crews to make it happen are actually on the ground.
Cost, payroll, and the paperwork that protects you
Hours become dollars when you apply a rate, and rates aren't uniform. An apprentice hour and a journeyman hour cost different money, and if you're costing everything at a blended average you're lying to yourself about where the labor dollars go. Track the categories separately or your cost variance reports are noise.
Two places this pays off directly. First, invoice verification: when a sub bills 1,200 hours for a period and your presence log shows 950, you've got a conversation to have, backed by data, before you cut the check. That single habit pays for the tracking effort several times over on a big job.
Second, if you're on public work with prevailing wage, activity-level hour tracking is most of the battle for certified payroll. You need to show who worked, in what classification, on what, for how long. Capture it clean in the field and the compliance paperwork is a byproduct instead of a monthly fire drill. Fake or sloppy certified payroll is how contractors lose the ability to bid public work — this is not the place to freelance.
The stuff that gets people hurt or thrown off the job
Two more things labor tracking quietly protects: certifications and safety. If an activity requires a certified welder, a licensed operator, or a specific OSHA card, "we're pretty sure Danny's certified" is not an answer that holds up after an incident. Verify credentials against the assignment. The same discipline that connects hours to activities lets you connect qualified people to the activities that require them, and flags the ones that aren't covered before the work starts, not after.
Safety-hour tracking is the unglamorous denominator behind every incident rate you report. Total man-hours worked is what your EMR and your recordable rates are calculated against, and if that number is soft, so is your safety story when an owner or an insurer asks for it. It also gives you a hard gate on orientation: nobody's hours count until they've been through your site-specific onboarding. A worker who never got the orientation is a worker your safety program can't account for, and that's a bad place to be standing when something goes wrong.
Don't drown in data — pick the three numbers that matter
You can measure everything and manage nothing. The trap with any labor tracking system is collecting a mountain of data that nobody reads. Resist it. On a running job, three views earn their keep: planned versus actual headcount by trade (are the bodies here?), hours-against-budget by activity (are we going to make the number?), and overtime by sub (what's quietly going sideways?). Everything else is analysis you do at milestones, not every morning.
The trends worth pulling at the end of a phase are the ones you can't see day to day — productivity drifting down over the course of a floor, output tanking on the weeks the weather turned, one sub consistently needing more hours than the others for the same scope. Those patterns are how you win the next bid instead of just surviving this one.
None of this requires a fancy platform to start. A clipboard and a foreman who counts heads honestly beats expensive software nobody updates. But once you're pairing daily labor with a live look-ahead — seeing the shortfall land on the activity it threatens, in time to move a crew or lean on a sub — you stop reacting to last week's problems and start managing next week's. That's the whole game. The wall tells you what's supposed to happen; the labor numbers tell you the truth about whether it will.