Prepare the venue, people and records
A proximity technology pilot does not run itself. Between the moment beacons are mounted and the point at which you evaluate results, there is an operational phase where things drift, fail or behave differently from expectations. Pilot monitoring and adjustment is the discipline of catching those problems early and correcting them before they distort your findings.
Monitoring during a pilot is distinct from long-term fleet management. You are not yet optimising for a multi-year rollout; you are watching for conditions that would make the pilot data unreliable. A beacon with a loose mounting that swings on its fitting, a zone trigger that fires only when a visitor stands still rather than walks past, or a batch of devices shipped with the wrong advertising interval will all produce misleading results if left unchecked.
Adjustment means changing something in the physical deployment or its configuration in response to what you observe. That might be moving a beacon thirty centimetres to avoid a metal pillar, reducing transmit power on a device that is bleeding into an adjacent zone, or tightening the RSSI threshold after real visitors turn out to carry phones in different positions than your test devices. The key principle is that every change is logged, time-stamped and tied to a specific observation, so that when you later evaluate the pilot you can separate stable conditions from periods affected by mid-pilot tweaks.
What Monitoring Covers
- Signal health: RSSI readings at defined test points, compared against your baseline measurements from installation day.
- Hardware status: battery voltage reports, firmware version confirmation, physical tamper checks.
- Trigger accuracy: whether zone entries and exits fire at the correct physical locations for real visitors, not just your test phone.
- Consent and engagement rates: the proportion of visitors who opt in and receive at least one notification, watched for sudden drops that suggest a technical fault rather than visitor disinterest.
- Environmental changes: new stock displays, moved furniture, closed doors or construction work that alters the radio environment.
None of this requires a bespoke dashboard in most cases. A structured logbook, a shared spreadsheet with timestamped entries and a routine walk of the pilot area with a monitoring app will cover a small-to-medium pilot. Larger deployments may justify automated alerting, but the discipline of regular physical checks remains valuable regardless of scale.
Pilot, measure and correct
Retail Entrance Zone Pilot
In a retail pilot, a common early finding is that the entrance beacon triggers too late or too early. During installation you may have tested with a phone held in your hand at chest height. Real visitors approach with phones in pockets, bags or pram organisers, and the antenna orientation and body absorption differ markedly. Within the first two days of a live pilot, you will typically see a cluster of triggers either metres inside the store or, conversely, not until the visitor is well past the welcome message zone.
The adjustment here is usually a combination of moving the beacon to a position with a clearer line of sight to the approach path and tightening or loosening the RSSI threshold in the receiving application. Because you are logging changes, the evaluation phase can treat the first 48 hours as a calibration period and analyse only the stable configuration that follows.
Museum Gallery Zone Pilot
Museum pilots often reveal interference from exhibit materials. A beacon mounted on a display case with a large metal frame or a glass panel with a metallic coating may broadcast reliably in one direction but be heavily attenuated in another. Visitors approaching from the weak-signal side either receive no trigger or receive it inconsistently.
Monitoring catches this when you compare trigger rates by approach direction. If your analytics show that visitors entering the gallery from the left trigger at twice the rate of those entering from the right, and the physical layout is otherwise symmetrical, the beacon placement relative to the case is the likely cause. Adjustment might involve repositioning the beacon to the top of the case rather than the side, or adding a second device to cover the attenuated direction.
Event Temporary Infrastructure
At events, the radio environment changes by the hour. A seminar room that is empty during setup fills with hundreds of bodies and laptops during a session, each absorbing and reflecting Bluetooth signals. A beacon calibrated in the empty room may under-trigger once the room is occupied.
Event pilots benefit from short monitoring cycles. Checking signal readings before doors open, during the first session and again at peak capacity gives you a realistic picture of the range you can expect. If the drop is severe, the practical adjustment is often to increase transmit power for the duration of the session and revert afterwards to conserve battery, rather than attempting to remount hardware mid-event.
Adjustment Workflow
A practical workflow for any adjustment during a pilot follows four steps:
- Observe: Record the symptom with specifics — time, location, device identifier, RSSI reading, visitor behaviour if visible.
- Diagnose: Rule out transient causes (a crowd momentarily blocking the signal, a phone in aeroplane mode) before attributing the issue to placement or configuration.
- Adjust: Make one change at a time. Moving a beacon and changing its transmit power in the same step makes it impossible to know which fixed the problem.
- Log: Record what you changed, when, and why. Include a note on whether the change is intended to be permanent or is a temporary experiment.
Document the result and remaining limits
Adjusting Without a Baseline
The most frequent mistake is making adjustments without a reliable baseline. If you did not record RSSI readings at defined test points during installation, you have no reference for whether a reading is normal or degraded. Before any adjustment, establish that baseline if it was missed. Walk the pilot zone with a known phone, record readings at fixed positions, and save them. Only then can you measure whether a change improved things.
Chasing Every Anomaly
Not every odd reading warrants an adjustment. Bluetooth signals fluctuate. A single visitor who reports not receiving a notification while five others did may have Bluetooth disabled, location services off, or a phone model with a known BLE receiver weakness. Distinguish between systematic patterns — the same zone consistently under-performing across multiple visitors and devices — and one-off anomalies that are part of normal variance.
Changing Too Many Variables
Altering beacon placement, advertising interval, transmit power and notification copy all within the same day creates a deployment where nobody can tell which variable influenced the results. Limit yourself to one category of change per monitoring cycle. If you move hardware on Monday, do not change software thresholds until Tuesday at the earliest.
Ignoring Staff Feedback
Floor staff often notice problems before your monitoring tools do. A retail colleague who mentions that visitors are confused by a notification that seems to belong to the wrong department is giving you zone-bleed information that RSSI logs may not clearly show. Build a simple feedback channel into the pilot — a shared messaging group or a short end-of-shift form — and treat it as a monitoring input alongside the technical data.
Forgetting Battery Drain Under Adjustment
Increasing transmit power or reducing the advertising interval to fix a coverage gap will shorten battery life. For a short pilot this may be acceptable, but if the adjusted configuration is intended to represent your production settings, the battery impact must be factored in. Check the manufacturer’s documentation for the adjusted parameters and recalculate expected life. If the pilot is meant to run for several weeks and your adjustment cuts battery life below that window, you need a different solution — additional beacons rather than higher power from fewer devices, for example.
Key Checks During a Pilot
- Daily: Review overnight trigger logs for gaps or spikes that suggest a beacon failure or a configuration error.
- Every two to three days: Walk the pilot zone with a monitoring device and compare live RSSI to the baseline.
- Weekly: Check battery voltage reports if the beacon platform provides them, and physically inspect mountings for drift or tampering.
- After any environmental change: If the venue moves a display, closes a partition or changes lighting (some LED drivers produce broadband RF noise), recheck affected zones within 24 hours.
- Before any evaluation period: Confirm that no unlogged adjustments have been made since the last recorded change. If they have, note them and consider excluding the affected period from analysis.
Pilot monitoring and adjustment is unglamorous but essential. It is the difference between a pilot that produces data you can trust and one that produces a set of numbers shaped by hardware faults, unrecorded tweaks and environmental changes you never noticed. Treat it as a defined phase with its own discipline, and the evaluation that follows will rest on solid ground.
