How Accurate Is Geofencing for a Time Clock? Radius Sizing for Real Job Sites

01

The three parts of the accuracy question

When an owner asks how accurate geofencing is, they usually mean "will it stop the worker clocking in from their driveway without blocking the one who is standing on the slab." That is three separate questions.

The first is the phone. A GPS fix has an error, measured in meters, and the size of that error changes with how much sky the phone can see. The second is the radius. You pick it, and it is a policy decision about how far from the site's center a worker can be and still count as on site.

The third is the app. When the phone reports a fix too rough to trust, the app has to do something with it, and that choice decides whether the worker in a basement gets blocked.

02

How accurate a phone's GPS fix is

GPS.gov, read September 8, 2026, states that "GPS-enabled smartphones are typically accurate to within a 4.9 m (16 ft.) radius under open sky" and that "their accuracy worsens near buildings, bridges, and trees." That is the only accuracy number we cite.

Under open sky, a phone knows where it is to within about 5 meters, roughly the length of one parking space. On a roof, in a field, or on a slab with nothing around it, a geofence works about as well as it ever will. Move the same phone next to a steel building, under a bridge deck, into a tree line, or down a stairwell, and the fix gets worse. GPS.gov does not put a number on how much worse, and neither will we, because it varies by the hour and by the site.

The error grows in exactly the places where crews work. No app can improve the reading; the app only decides what to do with it. Two consequences follow. A radius should be bigger than the error you expect at that site, or you get false blocks. And a site where the sky is mostly blocked needs a second check that does not depend on GPS.

03

Radius sizing by site type

The radius is a policy number, and you should set it per site rather than once for the company. A tight radius fits a residential call; the same radius on a 40-acre grading job blocks workers who are on site but far from the center. Start generous and tighten later, once you have seen a few weeks of clock-ins on the approval screen and know how far off the readings run at that site.

As an outside reference point, Intuit's help article on geofencing in QuickBooks Time, read September 8, 2026, says "The default and minimum radius size is 100 meters and can go up to 1000 meters." That range brackets the table above.

Put the center point where people stand when they clock in. A commercial address is often the front door, and the crew clocks in at the dock on the far side. A pin dropped on the mailing address can leave the dock near the edge of the fence.

04

False blocks and how to handle them

A false block is the case owners worry about: the worker is on site, the phone reports a position 40 meters outside the fence, and the punch is stopped. It happens most where the fix is worst, next to buildings, under trees, and indoors, and it is the reason the docs say to start generous. The hours are owed either way, so the procedure is about getting the record right.

1. 1The worker walks toward the center of the site, waits a moment for a better fix, and taps clock in again.
2. 2If the punch still fails, the worker tells the foreman the time they started and keeps working.
3. 3After the shift, the worker requests a correction from the app with the exact start minute and a reason, or the manager adds the shift from the web.
4. 4The manager reviews the request, checks the distance readout on later punches from that site, and approves.
5. 5If false blocks repeat at one site, the manager widens that site's radius or turns on the WiFi check for it.

The approval screen is where you learn how accurate your fences are. Every pending shift with a clock-in location shows a Map link and a readout such as "On site, 40 m from Main St job" or "Off the worksite, 310 m from Main St job." After two weeks you know how far from center the readings at each site usually land, and you size each fence to match. The approval flow is documented in approving timesheets.

05

What happens when the fix is too coarse to trust

There is a third case between inside and outside: the phone returns a position, but with an error so wide that inside and outside are both possible. A phone that has just come out of a basement, or one that has not yet found satellites, can report this kind of fix.

An app has two choices here. It can treat a wide fix as outside and block, which punishes the worker in the stairwell. Or it can treat the fix as unknown and let a person decide. Clox takes the second path: when a phone's GPS fix is too coarse to trust, Clox allows the punch and marks it for review. A manager reviews it with the rest of the week, where every pending shift with a clock-in location carries the Map link and the distance readout.

This is the only case where a geofenced punch goes through from an uncertain position. A geofenced worker whose phone reports a good fix outside every assigned worksite is blocked.

06

What this looks like in Clox

The radius lives on each worksite. Open the worksite, and the number you adjust is the radius around its center point. The worksites and geofence docs say the 200-meter default covers most single sites and that large commercial lots or sites with parking may want more. Geofencing itself is set per person on the Team page, under Worksites, with a toggle called Enforce geofence at clock-in and a list of the worksites that person may use. A roaming supervisor can be left off while the crew on the slab is on, and crews can be assigned to specific worksites.

The second check is the site WiFi network. Require the phone to be connected to the site's WiFi at clock-in, and the punch passes the location check without GPS. It is meant for indoor and multi-floor sites where GPS drifts, and it works on its own or together with the radius. Location is read only at the moment of the punch; Clox never tracks anyone between punches and has no live map. Punches also work with no signal: the phone queues the punch and sends it when the connection returns, which we explain in time tracking that works offline with GPS.

For the plain mechanics of a fence, read How does geofencing time tracking work. Geofencing vs GPS time tracking explains why recording a location and acting on it at the punch are different features.

07

What a geofence proves

The error above is the phone's. A worker can also fake a location, because location comes from the worker's phone. A geofence stops the casual driveway clock-in and makes a worker who wants to cheat go to some effort to get around it. It does not prove anyone was anywhere. We wrote about spoofing and what a time clock can and cannot do about it in Can employees fake their GPS location.

Clox punch records are signed and hash-chained, and anyone can verify a record at getclox.com/verify. The signature proves that nobody altered the record after capture, and it says nothing about whether the GPS reading inside the record was true.

08

FAQ

The 14-day Clox trial needs no credit card, and the first paid subscription has a 30-day money-back guarantee. Start a free trial, set one worksite at 200 meters, and read the distance readouts on Friday before you tighten anything.