How Accurate Is Your GPS, Really? A Practical Primer

Your phone will happily display a latitude with seven decimal places, but the number of digits is only formatting. Under good conditions, a modern phone may locate itself within several metres; indoors or in dense urban areas, Wi-Fi, cell and multipath effects can make the fix much less accurate. Precision on the screen and accuracy on the ground are different things.

A precision ladder worth memorising

• 3 decimals ≈ 110 m of latitude — neighbourhood scale.
• 4 decimals ≈ 11 m — building or field-gate scale.
• 5 decimals ≈ 1.1 m of numeric latitude precision — finer than many consumer fixes are actually accurate.
• 6+ decimals can be useful for storage, but should not be interpreted as proof of centimetre-level measurement accuracy.

A practical storage choice is signed decimal degrees on WGS84 with five or six decimals, then convert at the display edge for whoever needs DMS or another grid reference. The map preview is just as important as the digits: it catches swapped coordinates and missing minus signs quickly.

The useful habit is simple: paste the coordinates, inspect the pin, and compare it with what you know about the source measurement. A plausible pin does not make a coordinate legally or survey-accurate, but it catches many everyday mistakes before they spread into a dataset.

Accuracy, precision and repeatability are different

Accuracy describes how close the reported point is to the true location. Precision describes how finely the number is expressed. Repeatability asks whether repeated measurements cluster together. A phone can report many decimals (high numeric precision) while still being several metres wrong (limited accuracy), and repeated fixes can wander around the true point depending on sky view and surrounding buildings.

Why cities and buildings are difficult

Satellite signals can reflect from glass and concrete before reaching the receiver, a problem known as multipath. Indoors, the device may rely more heavily on Wi-Fi and network positioning. Dense tree cover and narrow streets can also reduce the quality of the satellite geometry. That is why the same phone can produce a tight fix in an open field and a much larger uncertainty circle in a city canyon.

A practical quality check

• Wait a few seconds for the location fix to settle instead of trusting the first point immediately.
• Look at the reported accuracy radius when the browser or device exposes one.
• Compare several fixes rather than assuming one reading is perfect.
• Inspect the pin against recognizable map features before storing or sharing it.
• For legal, engineering or survey work, use appropriate professional GNSS/survey equipment and controls.

The accuracy circle matters more than the last decimal

Many devices expose an estimated horizontal accuracy with the location fix. If that value is 20 metres, interpreting the sixth decimal place as meaningful physical certainty makes no sense. The reported point is better imagined as the centre of an uncertainty area. That mental model is especially useful when deciding whether a coordinate is good enough for a trailhead, a parcel corner or a building entrance.

Consumer devices can still be excellent for everyday navigation. The point is not that GPS is unreliable; it is that accuracy changes with conditions and the application should respect that variability.

Use Find My Location when you want the browser/device position, then compare the pin with known map features. If you need to understand why an IP lookup shows a different city, the guide GPS vs IP Address vs VPN explains the two location systems separately.

Why repeated measurements can help

If you record the same point several times, a cluster of fixes can reveal whether the device is stable or wandering. Averaging is not a magic substitute for professional surveying, but repeated observations can be more informative than trusting one instant reading. If all fixes jump in the same direction because of multipath or poor satellite geometry, averaging will not remove the underlying bias.

GPS altitude deserves extra skepticism

Vertical positioning is often less accurate than horizontal positioning, and different apps may also report height relative to different reference surfaces. A phone elevation number should not automatically be treated as equivalent to a surveyed orthometric height. For hiking and general orientation it may be useful; for engineering, flood design or construction it is not enough by itself.

When you need terrain context rather than raw device altitude, compare the coordinate with the Elevation Finder, which uses mapped elevation data and therefore has a different set of limitations from the phone sensor.

Privacy is part of the accuracy decision

Precise location is sensitive information. A site should ask permission before browser geolocation and should still provide a useful alternative when permission is denied. For many tasks, searching an address or clicking a map is enough. You do not need to reveal your live device position just to calculate a distance to a place you already know.

Match the tool to the consequence of being wrong

• Meeting a friend at a park: ordinary phone GPS is usually more than adequate.
• Recording a hiking waypoint: useful, but keep the map context and expect a few metres of uncertainty.
• Mapping utilities, property boundaries or structural work: use the professional positioning method required by the project.
• Emergency response: rely on the official emergency/location systems and procedures, not a consumer web page alone.

The best location workflow is not the one with the most decimal places. It is the one where the measurement quality is appropriate for the consequence of a mistake.