It is Tuesday morning and a property manager emails you: the third-floor restrooms were not touched last night. Your timesheet says two cleaners were on site from 6:15 p.m. to 9:40 p.m. You have no way to prove they were in the building, let alone on the third floor.
That gap between what your timesheet claims and what you can actually demonstrate is the reason geofencing showed up in janitorial software. It is not a productivity gadget. It is a way to attach a verified place to every recorded hour of labor.
Most explanations of the concept stop at "a virtual boundary around a job site." That is technically true and operationally useless. This article covers the mechanics, the accuracy limits you need to design around, how to pick a radius for each type of building you clean, and the failure modes that turn a good idea into a payroll dispute.
What geofencing actually is in a janitorial context
A geofence is a circle (occasionally a polygon) defined by a latitude, a longitude, and a radius. In cleaning operations it does one job: it decides whether a device requesting a clock-in is close enough to the job site for that clock-in to be considered valid.
The sequence is simple. A cleaner opens the timekeeping page on a phone. The browser or app asks the operating system for the device's current coordinates. The system compares those coordinates to the site's center point, calculates the straight-line distance, and either accepts the punch, flags it, or blocks it.
That comparison happens at a single moment in time. This is the single most misunderstood part of the technology, and it matters enormously for how you use it.
| Term | What it means in practice |
|---|---|
| Geofence center | The latitude and longitude you designate for the site. Usually the building entrance the crew actually uses, not the mailing address pin. |
| Radius | Distance from the center point in meters or feet. Any device inside this circle is treated as "on site." |
| Geofenced clock-in | A location check performed at the moment a punch is submitted. Passes or fails, then it is done. |
| Breadcrumb tracking | Continuous location logging while the shift is open. A different feature entirely, and far more invasive. |
| Geofence exception | A punch submitted from outside the radius, or with location unavailable. Needs a human decision, not silent rejection. |
| Drift | Normal variation in reported coordinates caused by satellite geometry, buildings, and weather. Not dishonesty. |
The accuracy reality you have to design around
GPS is not a surveying instrument in your cleaner's pocket. According to GPS.gov, GPS-enabled smartphones are typically accurate to within a 4.9 meter (16 foot) radius under open sky, and accuracy degrades near buildings, bridges, and trees.
"Under open sky" is the operative phrase. Your crews do not clock in under open sky. They clock in at a service entrance beside a six-story concrete wall, in an underground parking garage, or inside a windowless loading dock.
When the device cannot see enough satellites, it falls back to Wi-Fi positioning and cell tower triangulation. Those methods can be accurate to a few dozen meters in dense urban areas with good Wi-Fi mapping, and wildly imprecise in rural or industrial zones. A common operational pattern is that the same cleaner at the same door reports coordinates 10 meters off one night and 200 meters off the next.
How to size the radius for each site
A 150 meter radius, which is a common default, draws a circle roughly 300 meters (984 feet) across, covering about 17.5 acres. That is generous for a standalone suburban office and far too tight for a school campus.
Use this formula rather than guessing:
Radius = distance from center point to the farthest legitimate clock-in spot + 40 meter drift buffer
The farthest legitimate spot is usually the back corner of the employee parking area, the loading dock, or the far wing of a multi-building site. The buffer absorbs normal GPS error so honest cleaners are not flagged.
| Site type | Suggested radius | Reasoning |
|---|---|---|
| Small office suite in a shared building | 75 to 100 m (246 to 328 ft) | Tight footprint, but urban drift needs headroom. Cannot distinguish floors or tenants. |
| Standalone office, 20,000 to 60,000 sq ft | 150 m (492 ft) | Covers building plus surrounding parking. The default works well here. |
| Medical clinic or dental office | 100 to 150 m (328 to 492 ft) | Small buildings, often in medical parks where neighbors are close. Keep it tighter to avoid overlap. |
| Strip mall or retail plaza | 150 to 200 m (492 to 656 ft) | Long linear buildings. Measure from the middle of your unit, not the plaza address. |
| K-12 school or multi-building campus | 300 to 500 m (984 to 1,640 ft) | Detached gyms, portables, and bus lots all count as on site. |
| Warehouse or distribution center | 250 to 400 m (820 to 1,312 ft) | Truck yards and trailer lots push legitimate parking far from the office entrance. |
| High-rise with multiple tenants you serve | 100 to 150 m (328 to 492 ft) | One geofence cannot separate floors. Use separate work orders or checklists per tenant instead. |
| Industrial or rural site, poor cell coverage | 400 m (1,312 ft) or more | Wi-Fi and tower fallback is unreliable. Widen the fence or accept manual approvals. |
If you know the parcel size, you can sanity-check the radius. A roughly circular 5 acre site needs about 80 meters. A 10 acre site needs about 115 meters. A 40 acre campus needs about 225 meters before you add any buffer.
What geofencing proves, and what it does not
This distinction protects you from overselling accountability to a client and from underusing the data internally.
| Geofencing gives you | Geofencing does not give you |
|---|---|
| Proof a device was within X meters of the site at the punch timestamp | Proof the person holding the phone was your assigned cleaner |
| An objective start and stop time for payroll | Proof any task on the scope of work was completed |
| Evidence that ends most "we were there, we swear" disputes | Which floor, wing, or tenant space the crew occupied |
| Early detection of chronic late starts and short shifts | Movement between clock-in and clock-out, unless you separately enable continuous tracking |
| A defensible time record for wage and hour purposes | Detection of deliberately falsified device location |
Because location alone says nothing about quality, geofenced time should always sit next to a second layer of evidence: a checklist completed on site, timestamped photos of the restrooms and entry glass, or a periodic inspection score. Time plus verified work is what a client actually wants to see.
Implementation: a rollout that does not blow up in week one
1. Fix your site coordinates before anything else
Do not trust the geocoded pin from a mailing address. In office parks and rural areas, an address pin can land on the road centerline hundreds of feet from the building.
Set the center point on the door the crew enters. For multi-building campuses, set it on the geometric middle of the area they cover, not on the main office.
2. Write the policy before you turn the feature on
Your policy needs to state what is collected (a single location reading at clock-in and clock-out), what is not collected, who can see it, how long it is retained, and what happens when a punch fails.
Have every employee sign an acknowledgment. Several states, including Connecticut and New York, have employee notice requirements for electronic monitoring, and location data raises questions your attorney should answer for your specific states.
3. Decide now: block or flag?
Hard-blocking out-of-range punches feels decisive and creates wage and hour risk. Under the FLSA, you must pay for all hours worked and keep accurate records of them (29 CFR Part 516). A cleaner who worked four hours but could not clock in still gets paid four hours.
For most operators, flagging is the better default: allow the punch, mark it as an exception, and require a supervisor to review it before payroll closes. If you do block, you need a same-night fallback that is reliable, such as a text to the supervisor who then enters the punch manually.
4. Pilot on three sites for two weeks
Pick your easiest site, your hardest site, and one in the middle. The hard one is usually a basement, a parking garage, or a rural industrial building.
Track how many punches fall outside the radius and why. Adjust radii per location. Do not roll out company-wide until the exception rate on your pilot sites is low enough that a supervisor can review every exception in a few minutes.
5. Train on the phone, not in a meeting
Most failures are permission failures. If a cleaner denied the location prompt once, the browser remembers that choice and the punch silently fails forever after.
Walk each crew member through the actual prompt on their actual phone. Confirm one successful punch each. Ten minutes per person on the front end eliminates most of the support calls.
6. Run parallel for one pay period
Keep your old method, whether that is call-in, paper, or a spreadsheet, running alongside the geofenced system for one full pay cycle. Compare the two. The differences are the conversations you need to have with supervisors, and occasionally the reason you find a site that has been billed for hours nobody worked.
Pre-launch checklist
- Center coordinates set at the crew's actual entry door for every location
- Radius calculated per site, not one global default
- Written location policy distributed and signed
- Attorney review completed for every state where you operate
- Block-versus-flag decision made and documented
- Exception review owner named, with a deadline before payroll close
- Fallback procedure for dead phones, denied permissions, and no signal
- Every crew member has completed one successful test punch
- Supervisors trained to read the exception report
- Retention period for location records defined
Common mistakes that undermine geofencing
- One radius for the whole portfolio: A 150 meter default applied to a 60 acre school campus and a downtown suite at once guarantees both false failures and useless coverage.
- Setting the radius to punish rather than to measure: A 25 meter fence will flag honest cleaners nightly. Once supervisors start approving every exception without reading it, your data is worthless.
- Treating an out-of-range punch as theft: The most common causes are drift, an old phone, a denied permission, and a building that blocks signal. Investigate patterns across weeks, never a single incident.
- Nobody owns the exception report: If flagged punches are not reviewed before payroll closes, you have added a step and gained nothing.
- Promising clients more than the technology delivers: Never tell a property manager you can prove which floor was cleaned from a clock-in record. You cannot. Show them inspection photos instead.
- Ignoring the personal device question: Reimbursement rules for employee phone use vary by state. California, for example, requires reimbursement of necessary business expenses under Labor Code 2802. Sort this out before rollout, not after a complaint.
- Assuming location data cannot be faked: Developer tools and mock-location settings exist. Most systems, including browser-based ones, do not detect spoofing. Chronic anomalies still require supervision.
- Silent rollout: Crews who learn about location tracking from a failed punch at 6 p.m. on a Monday will assume the worst. Announce it, explain it, and explain what it does not do.
How often to review your geofence settings
| Frequency | What to review |
|---|---|
| Daily | Exception queue. Approve, correct, or investigate every flagged punch before it reaches payroll. |
| Weekly | Sites generating more than one exception per week. Usually a radius or coordinate problem, not a people problem. |
| Monthly | Punctuality and shift-length patterns by employee and by site. Compare clocked hours to your labor budget for each contract. |
| Quarterly | Radius audit across the portfolio. Rounding neutrality check. Confirm new hires signed the policy. |
| At every new contract | Set coordinates and radius during site walkthrough, before the first shift, using the same two-reading test. |
| Annually | Policy language, retention period, state law changes, and whether the data is actually changing decisions. |
Troubleshooting the four failures you will actually see
Location permission denied. The fix is in browser settings for that specific site, not in your software. Keep a one-page instruction sheet for iOS Safari and Android Chrome and text it to crews.
Punch lands 200 meters off in a basement or garage. Move the clock-in point to the door before entry, or widen the radius for that site. This is physics, not behavior.
Dead battery at end of shift. Missing clock-outs are the most common gap. Set a policy: the cleaner texts the supervisor, the supervisor enters the punch with a note the same night. Never leave it to memory at payroll time.
Two crews, one address. Multi-tenant towers and campuses need separate work orders or checklists per space, because a geofence cannot tell floor 3 from floor 11. Presence verification and scope verification are different tools.
How CleanTrack360 handles this
CleanTrack360 includes geofenced GPS clock-in and clock-out that runs in the phone browser, so crews do not need to install anything. The default radius is 150 meters and is configurable per location, which is what lets you set a tighter fence on a downtown suite and a wider one on a warehouse yard. Location is captured at clock-in and clock-out only, with no tracking in between, which is exactly the boundary you can put in writing for your team.
Because presence alone never answers a quality complaint, the same platform pairs those time records with quality inspections that use custom checklists, photo evidence, and automatic scoring, plus a browser-based client dashboard where property managers can see schedules and inspection reports themselves. Plans start at $99 per month for up to 5 team members, $199 for up to 20, and $249 for up to 50, priced per plan rather than per user, with a 14-day free trial and no credit card required.