Choosing a construction time clock app as a commercial sub: the questions that actually matter

Time sheets are the one piece of paper nobody defends. They arrive Monday, half of them in pencil, one of them in a truck in Waco. Somebody in the office rekeys them, somebody else asks why the Fairview job shows 240 hours when the foreman only had six guys, and payroll goes out anyway because it has to.

So you go looking for a time clock app. And here is the trap: almost every time clock app on the market was built to prove attendance. Did this person show up, where were they standing, how long were they there. That is the right question for a plumbing service company with 20 vans and the wrong question for you.

You are not trying to catch anybody. You are trying to make sure that when a crew of nine spends Tuesday on two different jobs, Tuesday’s cost lands on the two jobs in the right proportion — this week, not after month-end close.

That difference changes which features matter. Here is what to ask.

First, decide who is going to enter the hours

This is the decision everything else hangs on, and most buyers skip straight past it to feature lists.

Option one: every worker clocks themselves in and out. Clean data, timestamped, per person. It also requires that every worker has a phone they are willing to use for work, can be bothered at 6:28 a.m., and is somewhere with signal. On a high-rise core, a basement mechanical room, or a hospital addition where the GC takes phones at the gate, at least one of those is false.

Option two: the foreman enters the crew. One person, one screen, nine names, the job, the hours. It is less precise — the whole crew gets 6:30 to 3:00 whether or not two guys rolled in at 6:40 — and for job costing it is usually good enough, because job cost cares about the eight hours on the Fairview job far more than the eight minutes.

Most commercial subs end up on option two, or on both: workers clock themselves where it works, foremen enter the crew where it doesn’t. If an app only supports one of those, it has made your decision for you. Ask to see a foreman enter nine people against two jobs in one screen, on a phone, with one hand, standing up.

The eleven questions

Ask these of every vendor, including us. The answers sort the field-service apps from the ones that can carry a commercial crew.

1. Can one person enter time for a whole crew, in one pass? See it done. Count the taps.

2. Can a worker be on two or three jobs in one day? Punch work in the morning, the new job in the afternoon, an hour at the shop loading. If the app assumes one shift equals one job, every split day becomes an office correction.

3. What happens with no signal? Ask specifically: does the app hold entries on the phone and send them when signal comes back, or does the foreman lose the screen he just filled in? This is the question that most often has a disappointing answer, and the only way to know is to make them show you with airplane mode on.

4. What grain do the hours land at — job, or job plus cost code or phase? Be honest with yourself about what your foremen will actually split. Four codes they use beats forty they guess at. We wrote the whole argument up in tracking labor hours by job and cost code: pick your grain, then stop.

5. Who can edit an entry after the fact, and does the edit leave a trail? Hours get corrected — a foreman fat-fingers 10 instead of 1.0, somebody gets coded to the wrong job. That is normal. What is not acceptable is a system where last week’s numbers quietly change and nobody can say who changed them. If you ever have a labor dispute, a prevailing-wage audit, or a delay claim, the audit trail is the whole value of the record.

6. What rate does an hour get costed at, and when is that rate picked up? Rates change — a raise, an apprentice stepping up, a different fringe package. If the app recosts last March at today’s rate, your historical job costs are fiction. The hour should carry the rate that applied the day it was worked.

7. Is the costed hour the wage rate, or the burdened rate? These are different numbers, and the gap is large enough to change a margin decision. If a system costs at bare wage, your “labor cost” is missing workers’ comp, taxes, and whatever you carry on top — so know which one you are looking at before you make a margin decision on it. How to calculate your labor burden rate walks the arithmetic.

8. Do billable T&M hours and cost hours stay separate? They must. The hours you bill on a signed T&M tag are priced off a rate list with markup; the hours you cost are what the work actually consumed. Mixing them double-counts the extra and ruins both numbers. See how to price a T&M ticket.

9. What does it cost per field user? Ask the real number, for everyone you would ever want on it, including the two-week helper. Per-seat field pricing is a common reason a time clock app gets rolled out to four foremen and never to the crews — and a clock that half the field uses produces hours you still have to reconcile by hand.

10. Where do the hours go next? Into your payroll, into job cost, into both, or into a CSV somebody rekeys? “Exports to Excel” means a person in your office is still the integration.

11. Does it do anything with GPS, and can you turn that off? Some crews shrug at it; some take it as an accusation and the rollout dies in a week. Know what the app does, know what you can disable, and decide deliberately rather than finding out from a foreman.

The two features that get oversold

Geofencing. The pitch is that the app clocks people in automatically when they reach the site. On a commercial job the geofence covers the GC’s whole site, which includes the three other trades, the lay-down yard, and the sidewalk — and on a tower, everybody is at the same coordinates on sixteen floors. It is a decent convenience feature. It is not a cost control.

Facial recognition and biometric punch. Sold as a cure for buddy punching. Buddy punching is a real problem for a hundred-person hourly workforce with turnover. For a 25-person crew whose foreman knows exactly who showed up, the fraud you are preventing is rare and the friction you are adding is daily. Spend the money on getting hours onto the right job instead.

What usually breaks in the first month

• Nobody told the foremen why. If the message is “the office wants to track you,” you have lost. The message that works: this is how your extra work gets paid for, and how you stop getting asked about last week’s hours on a Friday afternoon.
• The job list is wrong. If the job a foreman needs isn’t in the picker, he picks the closest one, and now the hours are on the wrong job and look perfectly clean. Jobs have to be set up before the crew lands on them.
• Nobody owns Tuesday. Someone in the office has to look at yesterday’s entries daily, not weekly. Catching a miscoded day on Tuesday takes a minute; catching it at month-end takes an afternoon of archaeology.
• It arrived alone. The clock, the schedule, and the daily log are three views of the same day. If they live in three apps, your foreman does the same work three times, and the one with the least immediate payoff for him gets dropped. Hours are usually the one he drops.

How SubMark handles it

The time clock is part of the field side of SubMark, and the field side is free — foremen, crews, the time clock, daily logs, and every sub you invite to a job cost nothing, however many of them there are. Pricing is one company subscription with your first three office users included, then a per-seat price for each office user after that. That is deliberate: a clock that only the foremen use isn’t worth running.

What it does:

• Hours per worker, per job, per day, either clocked in and out or logged by a foreman, one entry per worker, so nine people don’t each have to do it themselves. Each hour keeps the rate that applied the day it was logged, so a raise doesn’t rewrite last March.
• Those hours are the hours job costing reads. Not an export, not a monthly reconciliation — the same record. Job costing shows labor at the wage rate, so you are looking at wage cost and should apply your own burden on top when you make a margin call.
• Crew scheduling and Where Workers Are sit next to it, so the plan for the week, who is on which job today, and the hours that got recorded are all one system. More on that in crew scheduling for commercial subcontractors.
• Daily logs from the same phone, which is where the narrative, the crew count and the photos go. See the foreman’s daily log.

Hours roll into weekly timesheets with worker signatures and supervisor approval, and a weekly payroll summary (hours per worker per day, straight time, overtime, gross) exports as a CSV for your payroll provider. The clock does not record location, so there is no GPS to turn off.

Some limits worth stating plainly. SubMark does not run your payroll and does not fill out or file certified payroll forms for you — the hours are the input your payroll runs from, not the paycheck. There is no direct payroll integration; the CSV is the handoff. Hours land on the job (and the phase, on jobs that use patch tracking), not on a cost-code list. The web app works on a phone, but there is no offline mode yet, so a basement with no signal is still a basement with no signal. And T&M hours are not job-cost hours: a T&M tag (an opt-in module) is priced in the office and signed on site as extra work, while cost hours come from the clock and the labor logs. That separation is intentional, and question 8 above is why.

The rest of the office side — change orders, the schedule of values, draws, job costing — is on the product page.

What to do this month

1. Decide who enters hours: workers, foremen, or both. Write it down.
2. Pick your cost-code grain. Four to eight codes a foreman will actually use.
3. Run one job on the new system for two weeks while paper keeps running alongside it. Compare the two at the end of week two.
4. Have one person in the office review yesterday’s hours every morning. Fifteen minutes.
5. Only then roll it out to every crew — and if field seats cost money, decide now whether you can afford the whole field, because half the field is the worst outcome.