Preparation and acceptance criteria
A site survey for a proximity or indoor navigation deployment means physically walking the space, measuring the radio environment, noting building materials, mapping obstacles and recording where power, mounting points and existing infrastructure sit. It is the step that turns a floor plan into a deployment plan. Skipping it means relying on assumptions about how Bluetooth signals, NFC reads or QR sightlines will behave in a space you have not properly measured.
Site surveys get dropped for understandable reasons. Project timelines are tight, the venue has already been photographed, the floor plan looks straightforward, or the integrator has done "similar buildings before." In some cases the decision is implicit rather than deliberate: a quick walk-through with a clipboard is treated as equivalent to a measured RF survey when it is not.
The practical consequence is that you lose the information needed to make three connected decisions: how many beacons or access points you actually need, where they should go, and what transmit power and advertising interval settings will produce reliable behaviour. Without those measurements, placement becomes guesswork, and guesswork in a radio environment is expensive to correct after installation.
A site survey also reveals problems that no floor plan can show. A heritage museum may have thick stone walls that attenuate Bluetooth far more than a modern retail unit's plasterboard partitions. An event venue may have metal staging, rigging or temporary structures that change the RF picture every time the layout shifts. A warehouse may have racking that blocks line-of-sight between beacons and creates dead zones that move when stock levels change. These are not edge cases; they are the normal conditions of real UK venues.
From preparation to live testing
Retail environments
In a standard high-street shop, the temptation to skip the survey is strongest because the space looks simple: open floor, predictable fixtures, a stockroom at the back. However, retail units often contain metal-framed shelving, refrigeration units, security barriers and mirrored surfaces that create unpredictable reflections and absorption. A survey would show whether a beacon intended to cover the footwear aisle is being shielded by a promotional display made of steel, or whether the stockroom door, when closed, cuts off the signal to a queue-detection zone. Without that check, you may discover the problem only when the campaign goes live and analytics show a gap in coverage.
Museums and heritage buildings
Heritage venues present the clearest case for a survey. Stone, brick and timber behave very differently from modern construction materials, and listed-building restrictions often limit where you can place hardware. A survey here is not just about radio propagation; it is about identifying viable mounting points that will not damage fabric, checking whether cases contain metal mesh that blocks NFC reads, and confirming that the lighting conditions allow QR codes to be scanned without glare. Skipping the survey in a museum typically leads to either under-deployment, where exhibits lack reliable coverage, or over-deployment, where beacons are clustered to compensate for unknown attenuation and battery replacement costs climb accordingly.
Event venues
Temporary events introduce a particular risk because the physical environment changes between setup and the live event. A survey conducted on an empty hall will not reflect conditions once staging, draping, AV rigs and crowd density are in place. The practical response is not to skip the survey but to schedule it at the right point in the build, ideally once major structures are in position but before the venue is fully occupied. Treating the empty-hall survey as sufficient is a common form of skipping by proxy.
Warehouses and industrial spaces
Large open-plan warehouses seem forgiving for radio signals, but racking systems filled with stock create corridors that channel or block Bluetooth depending on orientation. A survey in an empty warehouse will produce misleading results. If the survey is skipped entirely, the deployment may work during commissioning and fail once the facility is at normal stock levels.
Document the result and remaining limits
Mistaking a visual walk-through for a radio survey
Looking at a space and noting where things are is useful, but it does not replace measuring RSSI values at the positions where visitors' devices will actually be. A visual inspection will not tell you that a beacon three metres from a zone boundary is reading at the expected signal strength on one side of a partition and dropping below the detection threshold on the other. If you are going to shortcut the survey, be clear about what you are leaving unknown rather than assuming you have covered it.
Assuming all walls attenuate signals equally
A single brick wall, a glass partition, a plasterboard stud wall and a lift shaft all behave differently. Manufacturer documentation typically provides general guidance on material attenuation, but those figures are measured in controlled conditions. Your venue's specific construction, including any internal insulation, services routing or hidden metal, will shift the numbers. The only way to know the actual attenuation in your building is to measure it.
Ignoring the human body as an obstacle
At 2.4 GHz, the human body is a significant absorber of radio energy. A beacon that reads clearly in an empty corridor may show erratic RSSI values when the space is crowded. If your deployment is intended for peak periods, your survey measurements need to reflect that, or at least acknowledge the limitation. Skipping the survey means you will not have data to inform your calibration approach.
Ordering hardware before the survey
Purchasing beacons or access points based on a floor-plan estimate is a frequent consequence of skipping the survey. The result is either too few units, requiring a second order and delayed rollout, or too many, tying up budget in hardware that sits in a cupboard. A survey does not need to delay procurement entirely, but it should inform the final quantity. If you must order early, build a contingency into the numbers and treat the first installation as a calibration exercise rather than a final deployment.
Key checks to carry out instead of skipping entirely
If a full measured survey is genuinely not possible before deployment, there is a minimum set of checks that reduces the risk of a failed rollout:
- Walk every zone where beacons, NFC tags or QR codes will be placed and photograph the actual mounting positions, not just the general area.
- Note the building materials for every wall, partition and fixture between a beacon and its intended detection zone.
- Check for obvious sources of 2.4 GHz interference: Wi-Fi access points, microwaves, Bluetooth audio devices, industrial equipment.
- Confirm mounting method and whether it is physically possible at each planned location without drilling into protected surfaces.
- Verify line of sight for any QR placement, including lighting conditions and viewing angle from the visitor's position.
- Test at least one NFC tag read on the actual exhibit surface or case material, not on a sample in the office.
- Record the conditions under which you carried out these checks, so you can account for what was different when the space is occupied.
These checks do not replace a proper survey, but they provide enough information to make an informed decision about whether to proceed or to pause until a full survey can be completed. The mistake is not in doing less than a perfect survey; it is in proceeding as though the missing information does not matter.
If your deployment plan was drawn up without a site survey, the practical next step is to schedule one before fixing hardware to any surface. Even a single day of measured walking with a beacon and a scanning app will reveal whether your placement assumptions hold, and it is far cheaper than reworking an installation that was built on guesswork.
