Why distance estimates move
Accuracy in indoor positioning is not a fixed specification printed on a datasheet. It is a range of probable performance that shifts depending on the physical environment, the hardware involved, the number of people in the space and how the system has been calibrated. Any figure quoted before a site survey is, at best, a rough indication under ideal conditions.
The first practical step is to separate what a technology can achieve in a controlled test chamber from what it will deliver in your venue. Bluetooth beacons, for instance, broadcast a signal whose received strength (RSSI) is used to estimate distance. That estimate is affected by walls, shelving units, human bodies and even changes in stock levels. A museum gallery with stone walls and glass display cases will behave differently from a retail unit with metal racking, even if both use identical beacons set to the same transmit power.
Rather than asking a supplier for a single accuracy figure, a more useful approach is to define accuracy in terms of what your use case actually requires. If you need to trigger a notification when someone enters a footwear department, knowing they are within three to five metres of the zone boundary may be perfectly adequate. If you need to guide a visitor to a specific painting in a crowded gallery, the requirement is tighter and the infrastructure, calibration effort and cost will all increase accordingly.
Accuracy also varies by device. A high-end smartphone with a modern Bluetooth chipset will produce different RSSI readings from a budget model or a device in a thick case. If your visitors are using their own hardware, you are dealing with a heterogeneous mix of receivers, and your accuracy expectations need to account for that spread rather than assuming a single reference device.
Test the complete receiver path
Zone-level awareness
Many retail and venue applications do not need pinpoint accuracy. Knowing that a visitor has moved from the ground floor to the first floor, or from the entrance hall into the main exhibition space, can be achieved with a small number of beacons placed at transition points. The expectation here is not "you are two metres from the till" but "you are now in the homeware zone." This class of accuracy is broadly achievable with standard BLE beacons and minimal calibration, provided placement avoids obvious signal obstacles.
Exhibit or fixture-level triggering
Museums and heritage sites frequently want content to appear when a visitor is close to a specific exhibit. The practical question is how close is close enough, and what happens when two exhibits are near each other. If displays are spaced several metres apart, a well-calibrated beacon per exhibit can reliably distinguish between them. Where exhibits are clustered within a metre or two, the overlap in signal coverage makes clean separation difficult without additional techniques such as directional antennas, reduced transmit power or supplementary sensors.
Turn-by-turn wayfinding
The most demanding accuracy requirements come from turn-by-turn indoor navigation. A visitor following a route on their phone needs to know whether to turn left or right at a junction, and the system needs to detect that they have actually taken that turn. This typically requires a denser beacon grid, careful calibration at each decision point and often a fusion of Bluetooth with other inputs such as smartphone inertial sensors. Even then, accuracy will degrade in wide-open spaces where there are few reference points, and the expectation should be "reliable corridor-level guidance" rather than "sub-metre positioning at every step."
Events and temporary spaces
Pop-up venues and exhibition halls present a particular challenge because the environment changes rapidly. Stands are erected, removed and reconfigured between events. A system calibrated for one layout will not automatically perform the same way for the next. For event organisers, the practical expectation should be that each new floorplan requires at least a verification pass, and potentially a partial recalibration, rather than assuming the previous accuracy will carry over unchanged.
What changes invalidate the result
Taking manufacturer range figures at face value
A beacon datasheet may state a range of seventy metres or more. That figure typically refers to the maximum distance at which any signal can be detected in open air, not the distance at which reliable positioning occurs indoors. Using that number to plan beacon spacing is one of the most frequent planning errors. In practice, indoor placement is usually dictated by the need for overlapping coverage at a much shorter range, often between five and fifteen metres depending on the accuracy class required.
Calibrating once and assuming stability
Calibration involves measuring RSSI values at known distances in the actual venue and feeding those measurements into the positioning algorithm. The mistake is treating this as a one-time task. If you move stock, install new signage, add a temporary partition or simply see a change in footfall density, the RF environment shifts. A practical maintenance schedule should include periodic spot-checks at a sample of calibration points, particularly after any physical change to the space.
Ignoring the human body
The human body absorbs Bluetooth signals. A visitor standing between a beacon and their phone will produce a different reading from the same visitor holding the phone on the other side. In dense crowds, the cumulative effect of multiple bodies can reduce effective range noticeably. Any accuracy expectation that does not account for occupancy levels will be optimistic during busy periods.
Testing with the wrong devices
If your pilot uses a handful of company-issued smartphones, the results will not represent what your visitors experience. Test with a mix of device types, ages and case styles that reflect your actual audience. Record the spread of results, not just the average, because it is the outliers that cause missed triggers or incorrect zone assignments.
Key checks before committing to a deployment
- Ask the supplier whether their accuracy claims are based on lab tests or real-world deployments in environments similar to yours.
- Request a measured site survey rather than accepting a theoretical coverage plan.
- Define your minimum acceptable accuracy in terms of the use case, not as an abstract number.
- Agree on a verification method: how will you measure whether the system is meeting its accuracy target after installation?
- Clarify what happens to accuracy when the space changes layout, occupancy or contents.
- Check whether the positioning logic runs on the device, on a server, or both, and what the latency implications are for real-time wayfinding.
Setting accuracy expectations is fundamentally about aligning what the technology can realistically deliver in your specific environment with what your use case genuinely requires. Over-specifying accuracy drives unnecessary cost and complexity; under-specifying it leads to a system that frustrates visitors and fails to meet its purpose. The practical path is to define the coarsest accuracy that satisfies the use case, test whether the proposed infrastructure achieves it in your venue, and build in a process to verify that performance over time.
