The savings pitch for GPS clock-in almost always arrives as one tidy percentage of payroll recovered. That number is close to useless on its own, because it bundles deliberate time theft, which is rare on most crews, with rounding, drift and unapproved overtime, which is where nearly all of the money actually sits.
GPS time tracking reduces labor costs by replacing estimated hours with verified hours at a specific address. The savings show up in four places: early departures, late starts, favorable rounding, and overtime nobody approved. Measure the paid-to-scheduled hours ratio per site per week, and set your baseline before you change anything.
What follows is a set of benchmarks with their sources, the arithmetic that turns minutes into dollars, and the caveats that keep you out of a wage-and-hour problem. Treat every range here as a place to start looking, not a target to hit.
What is a good paid-to-scheduled hours ratio for a cleaning crew?
Paid-to-scheduled hours is the cleanest single measure of labor leakage in janitorial work. Divide total paid hours at a site by total scheduled hours at that site, weekly.
If you scheduled 60 hours and payroll shows 64.8, your ratio is 1.08. That gap is not automatically waste. Part of it is legitimate: a stripped floor, a client request, a cleaner who stayed to finish a restroom that got hammered. The point of measuring is that today you probably cannot tell which part is which.
Most operators who track this find the ratio varies wildly by site, not by employee. One account runs at 1.01 for months and another sits at 1.12 because the scope was never right in the first place. That distinction matters, because GPS clock-in fixes the first problem and only diagnoses the second.
Labor cost benchmarks for commercial cleaning crews
These are the numbers to check yourself against. Read the source column carefully: some rows come from published data, and some are ranges operators commonly use to flag outliers. The second kind is a screening tool, not a standard.
| Metric | How to calculate it | Starting range | Basis |
|---|---|---|---|
| Paid-to-scheduled hours ratio | Total paid hours ÷ total scheduled hours, per site, per week | 1.00 to 1.05 before you start questioning it | Operational screening range |
| Labor burden multiplier | (Wages + payroll taxes + workers comp + benefits + paid time off) ÷ wages | Compute your own. BLS reports benefits at roughly 30% of total compensation for private industry workers | BLS Employer Costs for Employee Compensation |
| Base hourly wage, janitors and cleaners | Look up your metro area, not the national figure | Varies by more than a factor of two between metros | BLS Occupational Employment and Wage Statistics, SOC 37-2011 |
| Loaded labor as a share of contract revenue | Loaded labor cost ÷ monthly contract value | Commonly managed in the 45% to 55% band on recurring janitorial | Operational screening range |
| Overtime as a share of paid hours | OT hours ÷ total paid hours, company-wide, monthly | Above 5% usually means a coverage problem, not a demand problem | Operational screening range; FLSA sets the 40-hour threshold |
| Shift start variance | Median minutes between scheduled start and actual clock-in | Within 10 minutes on established accounts | Operational screening range |
| Early departure rate | Shifts clocked out more than 15 minutes early ÷ total shifts | Measure your baseline first; there is no useful published figure | Measured, not benchmarked |
| Clock-in exception rate | Manual edits and missed punches ÷ total shifts | Under 5% after the first 60 days of rollout | Operational screening range |
| Production rate, sq ft per labor hour | Cleanable square feet ÷ labor hours on site | Build it from task times for your actual fixture and floor mix | ISSA, "540 Cleaning Times" |
| Consumer phone GPS accuracy, open sky | Manufacturer and system performance, not a setting you control | About 4.9 m (16 ft) radius; worse in urban canyons and inside structures | GPS.gov, "GPS Accuracy" |
How to build your loaded labor rate before you measure anything
Every dollar figure in this article depends on a loaded rate, not a base wage. If you price drift at $17.00 an hour when your true cost is closer to $20, you will underestimate the problem by nearly a fifth.
Here is an illustrative build. Every input is an assumption you must replace with your own numbers, especially the workers comp rate, which swings hard by state and by class code.
- Base wage: $17.00 per hour
- FICA at 7.65%: $1.30
- FUTA and SUTA, assumed 2.5%: $0.43
- Workers comp, assumed $4.00 per $100 of payroll: $0.68
- General liability allocation, assumed 1%: $0.17
- Paid time off accrual, assumed 2%: $0.34
That totals $19.92 per hour, a burden multiplier of about 1.17. FICA is a statutory rate. The rest are placeholders. Pull your actual workers comp declaration page and your state unemployment rate and rebuild this in ten minutes.
What six minutes a shift costs you across twelve sites
Take a 45,000 sq ft suburban office building cleaned five nights a week by three cleaners on four-hour shifts. Scheduled hours: 60 per week. Assume the loaded rate above, $19.92.
Now assume each cleaner leaves six minutes early, every shift. That is 0.1 hours across 15 shifts per week, or 1.5 hours. At $19.92 that is $29.88 a week and $1,554 a year at one building.
Six minutes is invisible. Nobody reports it, no supervisor writes it up, and the client never notices because the work still got done. It only becomes visible when the clock-out timestamp is tied to a location.
Scale it up. Suppose payroll at that same building shows 64.8 paid hours against 60 scheduled, a ratio of 1.08. The 4.8 excess hours cost $95.62 a week, or $4,972 a year. Across twelve buildings of similar size and behavior, that is roughly $59,660 annually.
You will not recover all of it, and you should not try. Some of those hours are real work you priced badly. If verified timekeeping moves twelve sites from 1.08 to 1.03, you recover 3.0 hours per site per week: $59.76 weekly, $3,107 per site per year, about $37,282 across the portfolio. Same assumptions, no invented efficiency claims, just arithmetic you can rerun with your own inputs.
How to measure your own baseline in 30 days
Do not roll out GPS clock-in and then try to calculate savings. You will have nothing to compare against and every number you report to yourself will be a guess.
Baseline worksheet, run this before you change anything
- Export four weeks of payroll hours by employee and by site.
- Export or rebuild the schedule those hours were supposed to match, in hours per site per week.
- Calculate the paid-to-scheduled ratio for every site. Sort descending.
- Flag any site above 1.05 and write one sentence next to it explaining why, from memory. Sites where you cannot write that sentence are your first targets.
- Count overtime hours and divide by total paid hours. Note which sites and which people generate it.
- Count timesheet edits, missed punches and paper corrections. This is your current exception rate and it will look better than reality.
- Build your loaded rate from your own tax and insurance documents.
- Multiply the excess hours at your five worst sites by the loaded rate. That is the size of the prize, and it is the only number worth quoting internally.
Keep that worksheet. Ninety days after rollout, rerun it identically. The comparison is the only honest measure of whether the change paid for itself.
What geofence radius should you set for a job site?
A geofence that is too tight generates false rejections, and false rejections generate manual overrides, which puts you right back where you started. Too loose and a cleaner can clock in from the parking lot across the street.
GPS.gov reports that consumer GPS receivers under open sky are accurate to roughly a 4.9 m radius. Inside a building, next to a curtain wall, or between two towers downtown, that degrades considerably. Your radius has to absorb the error, not fight it.
- Standalone suburban office or retail: a radius of roughly 100 to 150 m is usually generous enough to prevent false rejections at the door.
- Multi-building campus or medical park: set a fence per building, not one fence around the whole property, or you lose the ability to tell where anyone actually is.
- Dense urban high-rise: widen it. Signal reflection off surrounding buildings is real and the alternative is a nightly override queue.
- Strip centers with shared lots: a wide fence may overlap two of your accounts. Rely on the shift assignment, not the fence, to distinguish them.
Is GPS clock-in legal, and do you have to pay a blocked shift?
Location-verified timekeeping is lawful in most US jurisdictions when it applies during work hours and employees are told about it in writing. Some states have specific consent, notice or electronic monitoring statutes, so have your handbook language reviewed locally.
The rule that catches people is simpler. Under the FLSA you owe pay for hours worked, full stop. If the geofence rejects a clock-in and the cleaner works the shift anyway, that shift is compensable and your system needs a documented exception path to capture it.
Rounding is also governed. Under 29 CFR 785.48, rounding to the nearest 5, 10 or 15 minutes is permitted only if it works out neutrally over time. A system configured to round every clock-in up and every clock-out down is a back-pay claim waiting to be filed. If your current "savings" come from asymmetric rounding, that is not a saving, it is a liability.
One more, easy to miss: under 29 CFR 785.38, travel between job sites during the workday is compensable. If your crews hit three buildings a night and you only pay from first clock-in to last clock-out at each site, check that policy before you put timestamps on everything.
What GPS time tracking will not fix
Being honest about the limits is how you avoid buying a tool for a problem it does not solve.
- An underpriced contract: if you sold 45,000 sq ft for hours that were never enough, verified timekeeping just documents the shortfall more precisely.
- Work quality: a timestamp proves presence, not that the restrooms got done. That is what inspection scores are for.
- What happens between punches: most systems capture location at clock-in and clock-out only. A cleaner can punch in at the door and sit in the truck, and the record will look clean.
- Buddy punching entirely: location verification stops someone punching a coworker in from home. It does not stop two people who ride to the site together.
- Spoofed location: mock-location apps exist. Treat GPS data as strong evidence, not proof, and pair it with supervisor visits and inspection records.
The realistic value is narrower and still worth real money: verified start and stop times at a known address, an auditable record when a client disputes coverage, and a per-site ratio you can act on.
Questions operators ask before rolling out GPS clock-in
Can I require employees to use their personal phones for GPS clock-in?
In most states, yes, provided you disclose it in writing and only track during working time. California and a handful of other states require reimbursement of a reasonable share of business use of a personal phone under statutes like Labor Code 2802. Budget a monthly stipend if you operate there, and offer a company device to anyone who genuinely has no smartphone.
How long before payroll numbers actually change?
Behavior at the clock tends to settle within two payroll cycles, because the feedback is immediate and visible. Structural savings take longer, usually 60 to 90 days, since they require rebuilding schedules at sites where the data exposes a scope or staffing mismatch. Rerun your baseline worksheet at 90 days rather than reacting to the first clean week.
Does GPS clock-in stop buddy punching?
It stops the remote version, where one employee punches another in from a different location, and that is the most common form. It does not stop two people arriving together in one car when only one is working. Pair location data with unannounced supervisor visits and inspection scoring if a specific site keeps producing suspicious patterns.
Do I have to pay a cleaner who could not clock in because of the geofence?
Yes. The FLSA obligates you to pay for hours worked regardless of whether your timekeeping system cooperated. Publish a written exception procedure: the cleaner texts the supervisor at start and finish, the supervisor enters the time, and both note the reason. Then investigate the site, because repeated failures usually mean the radius is too tight or cell coverage is poor.
Should the geofence radius be the same at every location?
No. A radius that works at a freestanding suburban office will generate constant rejections at a downtown tower where signal bounces off surrounding buildings. Set a default, then tune per location using your override rate as the signal. Any site producing repeated manual entries needs a wider fence or a different approach, not more coaching.
Where CleanTrack360 fits
CleanTrack360 includes geofenced GPS clock-in and clock-out that runs in the phone browser, so crews do not install anything. The default radius is 150 m and is configurable per location, which is what you need when a downtown account and a suburban one behave differently. Location is captured at clock-in and clock-out only, not continuously between them. There is no published mobile app yet, the native app is still in development, and reports export to CSV so you can run the paid-to-scheduled math in your own spreadsheet.
Plans are priced per plan rather than per user: Starter at $99 per month for up to 5 team members, Pro at $199 for up to 20, and Business at $249 for up to 50. Scheduling, inspections with photo evidence, and a client dashboard sit alongside the time data, which matters because a timestamp tells you someone was there and an inspection score tells you what they did. There is a 14-day free trial with no credit card if you want to run your baseline worksheet against real data first.