How accurate is a Bluetooth tracker? The honest answer

Short answer: a coin-cell Bluetooth tracker is accurate to a room indoors and to a street or a building outdoors. It does not produce a metre-perfect dot, and anything that promises one is either using a different radio or exaggerating. If you want to know how accurate is a Bluetooth tracker before you spend money, that sentence is the whole of it — the rest of this piece explains where the error comes from and how to shrink it.

Where the position on your map actually comes from

There are two completely separate mechanisms, and they have different accuracies.

Inside Bluetooth range, your own phone is talking to the tag. It knows the tag is nearby because it can hear the broadcast, and it estimates how near from how strong that broadcast arrives. That estimate is coarse.

Outside Bluetooth range, the position is second-hand. Somebody else's Apple device walked past, heard the tag and reported an encrypted location back. The accuracy of that report is the accuracy of their phone's location at that moment, which in a city is usually good to a building and in open country can be much better.

So the map pin is not the tag saying where it is. The tag has no idea where it is. The pin is a record of where a phone was standing when it last overheard a broadcast.

The accuracy you should expect, scenario by scenario

The pattern is worth noticing: accuracy is not a property of the tag at all. It is a property of how many Apple devices happen to be near it. Somewhere with constant footfall — a station concourse, a high street, an open-plan office — the pin refreshes again and again. Put the identical tag in a rural outbuilding with nobody walking past and it can sit unchanged until morning.

Why the error exists: signal strength is a bad ruler

Bluetooth Low Energy works in the 2.4 GHz band — 2400 MHz to 2483.5 MHz, divided by the Bluetooth SIG's Low Energy primer into forty channels — and a receiver measures how loud an arriving packet is. Loud means close. Quiet means far. That relationship holds beautifully in an empty field and falls apart in a house.

Three things break it:

1. Absorption. Water absorbs 2.4 GHz, and a human body is mostly water. Stand between your phone and the tag and the reading drops as if the tag had moved several metres away.
2. Reflection. A signal bouncing off a radiator, a fridge and a metal door frame arrives by several paths at once. The receiver sees the sum, which can be louder or quieter than the direct path alone.
3. Orientation. A tag lying flat under a sofa cushion radiates differently from one hanging on a hook. Nothing moved; the number changed anyway.

This is the reason serious ranging needed a different approach altogether. Ultra Wideband times the flight of a pulse instead of measuring loudness, which is why AirTag can draw an arrow and a coin tag cannot. The Bluetooth SIG adopted its own answer, Channel Sounding, in Core Specification 6.0 on 3 September 2024, but it needs new silicon at both ends and is not in the tags on sale now.

Five ways to get better accuracy out of the tag you own

1. Hang it, do not bury it. A tag in an outside pocket of a rucksack reports far better than one under a laptop and a wet coat.
2. Keep metal off it. A steel toolbox, a car boot floor and a tin of screws all behave like a shield.
3. Use the sound, not the screen. Once you are in the right room, stop staring at the map and press the button.
4. Give the network time. A stale pin is not a wrong pin. Refresh, wait, and watch whether it moves.
5. Tag the bag, not the small thing inside it. Bigger objects sit in more open positions and get heard more often.

How BlueTag deals with the last few metres

BlueTag does not pretend to be a surveying instrument. Within about 60 m / 200 ft of your iPhone, with nothing much in the way, it holds a link and your phone can set the buzzer off at 80–100 dB, which is the part that genuinely locates something behind a cushion. Past that distance the Apple Find My network carries it, and every sighting arrives encrypted so only your Apple Account can read it.

Ultra Wideband is absent, so no arrow will ever appear, and we would rather write that here than let you discover it on a wet Tuesday. What there is: an IP67 case at 36.3 × 36.3 × 9 mm, a coin cell good for a couple of years, an iPhone or iPad from iOS 14.5 upwards, and a pair of tags per box for under £15, posted free. If the room-level answer is enough for the things you lose, see the specification.

Frequently asked

Does a Bluetooth tracker show live movement? Inside your own phone's range, effectively yes — the connection is continuous. Outside it, no. You see a series of sightings contributed by passing devices, so a moving bag updates in jumps rather than as a smooth line.

Why does the map sometimes show a place the tag has never been? Because the reporting phone's own location was poor. Deep indoors, a phone falls back on Wi-Fi and cell positioning, and the pin inherits that error. Treat a single odd sighting as noise and wait for the next one.

Is a more expensive tag more accurate? Only if the extra money buys a second radio. Paying more for a louder speaker or a nicer case changes nothing about position accuracy; paying more for Ultra Wideband does. Our technical comparison with GPS covers the other end of that scale.