Payroll losses in field work
How much value is hidden in unclear labor hours?
There's no credible universal figure for how much contractors lose on payroll, and the percentages repeated online don't trace back to a published study. What you can work out is your own number: the hours each week that get rounded, reconstructed from memory or charged to the wrong job, multiplied by crew size and loaded labor rate. The calculator below does that arithmetic in the open.
None of this assumes anyone is doing anything wrong. Most unclear hours come from paper, memory and crews moving between sites.
- Arithmetic you can check
- Your inputs, not an industry average
- Patterns, not accusations
Tue Oct 6
Paper sheet · typical entry
Harbor Yard · 6:30 AM – 4:00 PM
9.5 h
Rounded to the half hour. One job for the whole day.
Badge record · same day
- Shop & OfficeOn site25m
- Transit between locationsTransit1h 35m
- Riverside Medical PavilionOn site45m
- Cascade Supply Co.Vendor25m
- Harbor Yard Distribution CenterOn site6h 10m
Charged to Harbor Yard beyond the record
9.5 − 6.2 = 3.3 h
0.2 h of rounding; the rest was time at other locations or between them, now on its own lines.
The honest answer
Why there's no reliable "percent of payroll" figure for field crews.
The numbers that circulate (a share of payroll, a national dollar total) usually come from vendor blogs or from a hypothetical estimate that has since been quoted as fact. They also measure different things: rounding, data-entry errors, unpaid overtime, misallocated job costs. Your crews, sites and pay rules are your own, so the useful figure is the one you calculate from them. Errors also run both ways: unclear hours can mean someone was paid too little as well as too much.
Context: the US Department of Labor's Wage and Hour Division recovered more than $259 million in back wages for 176,957 workers across all industries in fiscal year 2025 (DOL WHD data, FY2025).
Estimate your own
Three hours a week, at your crew size and your rate.
Set two things: your field headcount and your average loaded labor cost per hour (wage plus taxes, insurance and benefits). One assumption stays fixed, 3 hours per worker per week across 50 working weeks, and the result is the labor value of those hours over a year, with the arithmetic shown.
Use it to size the question, not to book a saving. The point is to decide whether it's worth looking at your records, and where to start.
Labor cost of hours lost, per year
$168,750
Labor cost of hours lost per year: $168,750
25 workers × 3 h × 50 weeks × $45/h
- Hours lost / week
- 75
- Hours lost / year
- 3,750
Hours lost, building week by week3,750 h by week 50
That's the paid working year of about 1.9 full-time workers.
Worked arithmetic
Three crews, with the arithmetic shown.
Two cases are paid hours nobody can account for. The third is different: payroll is right, but the hours are charged to the wrong job, which is where job cost reports go quiet.
crew × minutes ÷ 60 × loaded rate × workdays
Assumed inputsAA mid-size mechanical crew
Start and finish times rounded or written from memory
25 people × 12 min ÷ 60 = 5 h a day × $45/h = $225 a day × 240 days
$54,000a year · 1,200 labor hours
BA larger multi-site contractor
Hand-entered punches and unrecorded time between sites
120 people × 8 min ÷ 60 = 16 h a day × $52/h = $832 a day × 245 days
$203,840a year · 3,920 labor hours
CForemen and leads who split their day
Time spent at a second job, coded to the first
8 people × 45 min ÷ 60 = 6 h a day × $76/h = $456 a day × 240 days
$109,440a year of labor charged to the wrong job · payroll unchanged
Where the value hides
Five places unclear hours hide on an ordinary timecard.
$39.69
average hourly wage in construction, before taxes, insurance and benefits are added.
Rounding
Start and finish written to the nearest quarter or half hour. Each instance is small; it repeats every shift.
Punches entered by hand
A late start or early finish written down as the scheduled time, often days later and in good faith.
Unallocated time between sites
Drives and supply runs that land on no job, or on every job, depending on who filled in the sheet.
Misallocated job codes
Hours paid correctly but charged to the wrong project, so one job looks cheaper than it was and another looks worse.
The office's own time
Hours spent chasing missing sheets, interpreting handwriting and re-keying the week into payroll.
Investigate your records
Run two pay periods side by side and compare.
You don't need a consultant. You need two pay periods and something to compare the timecards against.
- 01
Run a parallel period
Keep your current timecards for one or two pay periods while badges record the same shifts.2 pay periods
- 02
Compare day by day
Line up entered start, finish and job against the badge-recorded day. Note the differences by type.Office
- 03
Look for patterns, not people
Group differences by site, crew and kind. A consistent gap usually points to a process, not a person.By site
- 04
Fix the process
Realistic scheduled starts, job codes set by location, supply runs on their own line. Then measure again.Ongoing
ZeroBadge AI
Ask ZeroBadge AI to size the gap in your own records
The arithmetic on this page uses your inputs. Once the badges are running, ZeroBadge AI can answer the same questions from your real records, with every number linked to the days behind it. It is read-only and finds patterns, not culprits.
You ask
How many paid hours last month were not tied to a jobsite?
640 of 6,200 paid hours, about 10%, were transit, supplier stops or time with no location reports. 190 of those hours had no location reports at all, and each one links to its day.
You ask
How often did shifts start late or end early?
Late starts on 58 of 612 worker-days and early finishes on 23, so 81 worker-days or 13%. Approved time off is left out of the count.
You ask
Which crew has the most unverified time?
The Elm Street crew, with 94 hours against 31 for Harbor Road. Most of it is stretches with no location reports, which lowers the record's confidence rather than pointing at a person.
Answers come only from your company's own records. Every number links to the records behind it. Read-only; admins control access.
How ZeroBadge AI worksQuestions
Questions about unclear labor hours.
Is there an industry average for payroll leakage in construction or other field work?
Does this mean my crews are padding their hours?
How do I estimate the unclear hours for my own crew?
Will ZeroBadge reduce my labor costs?
Keep exploring
Next: labor cost by job, paper and payroll.
- SolutionLabor cost trackingLoaded rates, job budgets and forecasts from badge-recorded hours.
- SolutionReplace paper timesheetsTimesheets built from the badge record, so the office stops re-keying the week.
- SolutionMulti-jobsite time trackingSite transfers, transit and supply runs inside one shift record.
- SolutionPayroll integrationApproved hours sent to your payroll provider automatically. Nobody re-keys a timecard.
- Platform tourThe platformA walkthrough of what the office sees: crews, jobsites, timesheets, analytics and payroll.
- SolutionAll solutionsHow the badge, the platform and your payroll fit together.
Get started
Put your own numbers next to your own records.
We'll walk through a parallel pay period: what the badge records, what the timecards say, and where they differ.
What we will walk through
- 01Your crew size, rates and working days
- 02How differences are grouped by site and type
- 03What a parallel period looks like in practice