How it works
Let's refresh some of the basic... Spaces collect a large amount of presence/location signals from all the nearby devices and process them to extract and classify the meaningful interactions that represent actual store visits. In fact, not all the signals we collect are related to actual visitors, shoppers or clients, but comes from walk-by traffic, nearby stores or even other network devices; Spaces elaborate these data with intelligent algorithms that automatically separate relevant data from noise and turns it in actionable metrics.
It is important to highlight that the baseline tracking technology used to track these events, cannot detect deterministically all the devices that are in proximity of the sensing devices; in fact, many smartphones adopt measures to protect their identity using random identifiers and many sensing devices (for example Cisco Maraki) do not even track these devices in the first place. This is one of the reasons why we focused the solution on those analysis and use cases that are not affected by this phenomenon.
We removed some of the metrics that refer to the count of unique individuals. The metrics have been removed from the Cloud4Wi Dashboard (including Custom Reports), from the CSV export and from API endpoints.
In particular, we removed the following metrics:
- Visitors / Potential Visitors / Missed Visitors
- Visits / Potential Visits / Missed Visits
- Identified Visitors
- New Visitors Rate (complement of Returning Visitors Rate)
- Engagement Rate (complement of Bounce Rate)
We introduced a new metric called Traffic Index that is a proxy of the number of unique visitors, and it replaces the one previously called "Visitors". The new nomenclature reflects the fact that the metrics don't represent the exact count of individual that have visited a location.
The new "Behaviors" dashboard
The new dashboard includes multiple charts and metrics.
The top chart includes a history chart that plots the metrics across the observation period. In the history chart, each value represents the average value of the metrics for each time slot (for example a day) represented in the chart. The spot values on top of the chart, instead, represent the average value of the selected metric across the entire observation period.
The metrics in this chart are:
- Visit duration: this is the average duration of the visits.
The time visitors spend in the store is a good proxy of engagement. By tracking the history of this metric, brands can understand what store marketing initiatives are working and which ones don’t keep visitors engaged
- Bounce Rate: this metric represents the percentage of visits shorter than a threshold (customizable) minutes.
This metric, for example, helps measure the effectiveness of promotions displayed in the store entry area to capture visitor attention.
- Attraction Rate: this metric is a proxy of the attractiveness of the store, and it is calculated as the percentage of foot traffic detected in the proximity of the location that is converted into an actual visit.
By comparing this value across multiple stores, the client can, for example, identify what displays and window setups work better for attracting visitors.
- Returning Rate: a returning visit represent a visit done by an individual that already visited the brand before that specific event. The returning rate reported for a specific time slot is computed as the percentage of visits in that time slot that are classified as returning.
If a device does a visit on day 1 and then a second visit from the same device on day 2, the second visit is classified as "returning". Note that a visit is considered returning only if the previous one happened on a different day.
The Returning Rate is a proxy for customer loyalty and its evolution can help assess the effectiveness of retargeting campaigns and customer satisfaction.
The second chart is a histogram that shows the distribution of the Visit duration, to help you understand, other than the average value shown above, how the metric is actually distributed.
The last history charts represent the value of the Traffic Index. Traffic Index is a proxy for the store traffic. While it can't tell exactly how many visitors visited during a certain period, this metric will help you understand how traffic intensity changes over time or how different stores perform compared to each other. Traffic Index is still represented with a number to facilitate comparison in time and across the stores, but it actually does not represent a count of visits or visitors.
As a consequence of these changes, also the parameters for blacklisting staff and network devices have changed. For analytics and reporting purposes, you can now discard any visit longer than a configurable visit time.
If you are currently using the APIs to extract those metrics, you probably don't need to adapt your integrations. The existing API endpoint won't be removed and data model won't change, but they won't return any data for the metrics that will be removed.
Privacy and security
The signals collected from presence/location sources include the WiFi MAC address of the individual's phones, which is considered personal data under GDPR and other privacy regulations. Spaces immediately anonymize these identifiers in an irreversible fashion employing sophisticated processing, not using simple pseudonymization, but actually introducing a considerable level of information loss that makes mathematically impossible to know with a reasonable degree of certainty what the original client MAC address was. Moreover, for further robustness, Cloud4Wi also completely deletes the anonymized identifiers within 24 hours.