Define the decision before the metric
Proximity technology introduces metrics that behave differently to standard web analytics. A website records a page view with reasonable certainty; a Bluetooth beacon only records that a compatible smartphone with an active receiver was within range. Translating a radio signal into a business objective requires separating what the hardware detects from what the visitor actually does.
The first distinction to make is between detection metrics and action metrics. A detection metric might be the number of times a beacon’s identifier was logged by passing devices. An action metric is the number of times a visitor opened a notification, tapped an NFC tag, or completed a journey between two points. Objectives tied solely to detection are rarely useful, because a high signal count does not confirm attention or intent.
Objectives in physical spaces generally fall into two categories: operational and commercial. Operational objectives focus on how people move and wait—reducing queue dwell time, improving wayfinding completion rates, or balancing footfall across zones. Commercial objectives focus on transactions or defined conversions, such as redemption rates for location-triggered offers or upsell rates at specific exhibits. Attempting to measure both simultaneously in a single pilot often obscures the results, so it is usually better to prioritise one category.
Another fundamental consideration is the consent denominator. Under current UK privacy guidance, a visitor must opt in before their device can be tracked continuously. If your footfall is 1,000 people but only 200 have consented, your baseline for any percentage-based KPI is 200, not 1,000. Setting an objective without accounting for the likely consent rate will produce targets that are mathematically impossible to hit.
Bias, missing observations and denominators
Because the measurable outcomes change depending on the venue, objectives must be tied to a specific physical context rather than copied from another sector.
Retail environments
A common objective is to increase conversion rates at a specific fixture or till point. The practical KPI might be the percentage of opted-in visitors who received a notification near a product display and subsequently completed a purchase linked to that campaign. To make this measurable, the venue needs a reliable baseline: what was the conversion rate for that display before the beacons were installed? Without a control period or a control zone, attributing a change in sales to the proximity system rather than seasonal demand or merchandising changes is guesswork.
Queue management is another frequent objective. Here the KPI is not a sale, but time. The practical metric is the average dwell time in a defined queue zone before and after introducing a notification that redirects visitors to quieter tills. This requires accurate zone calibration, as the system must distinguish between someone browsing an endcap and someone actively waiting in a queue.
Museums and cultural venues
Objectives here usually centre on content engagement and visitor flow. A typical KPI is the content trigger rate: the proportion of visitors who arrive at an exhibit zone and actively request the associated audio or text content, either via an app prompt or an NFC tap. A secondary KPI might be dwell time per exhibit, measuring whether proximity-triggered content increases the time spent at a particular display compared to the pre-deployment average.
Flow balancing is an operational objective for popular venues. If a museum knows that Zone A becomes congested at 11:00, an objective might be to shift 15% of that footfall to Zone B during that window. The KPI is the change in detected device counts per zone during the target hour, measured against historical data for the same day of the week.
Events and conferences
Event objectives often focus on reducing friction. A practical KPI is the average time taken for an attendee to navigate from the registration desk to their first scheduled session, measured by the time between a check-in beacon detection and the first session-room beacon detection. Another objective might be reducing no-shows at breakout sessions by sending a reminder notification when the attendee is still in the catering area ten minutes before the session starts. The KPI here is the attendance rate for targeted attendees versus a non-targeted control group.
Retention, governance and ownership
The most frequent error is treating proximity KPIs like digital marketing KPIs. A two percent click-through rate might be acceptable for a display advert, but if only two percent of opted-in visitors interact with a beacon notification, the pilot is likely failing to deliver relevant content or the triggers are firing at the wrong moment. Physical interactions demand higher relevance thresholds because the visitor is already present and invested in the environment.
A second mistake is ignoring the calibration lag. During the first weeks of a deployment, RSSI values shift as staff adjust beacon placement, stock changes on shelves, or seasonal decorations alter the radio environment. KPIs measured during this settling period are unreliable. Objectives should include a stabilisation window—often two to four weeks—before formal measurement begins.
The third pitfall is setting KPIs that the operations team cannot directly influence. If the objective is to increase overall daily revenue by five percent, but the proximity system only covers one department, the operations team has no control over the other variables affecting that figure. KPIs must map to the specific zones and interactions the technology governs.
Before finalising an objective and KPI framework, run through these checks:
- Is the baseline documented? Ensure you have measured the current state—dwell times, conversion rates, or queue lengths—before any hardware is installed.
- Is the denominator defined? Be explicit about whether a percentage is calculated against total footfall, opted-in footfall, or devices that successfully received a signal.
- Is the KPI actionable? If the metric drops, identify exactly which lever the operations team can pull to improve it, such as adjusting transmit power, rewriting notification copy, or recalibrating a zone boundary.
- Does the objective respect privacy constraints? Verify that the data required to calculate the KPI can be collected lawfully under current UK guidance, and that retention periods are not extended simply to make a historical comparison easier.
- Is there a control? Wherever possible, designate a similar zone without the technology to distinguish the system’s impact from general environmental changes.
Setting rigorous objectives at the outset prevents the common trajectory of a proximity pilot: installing hardware, collecting large volumes of raw signal data, and then struggling to explain what any of it means for the business. By defining the measurable outcome first, every subsequent decision—from placement to calibration to notification frequency—can be tested against that specific target.
