Dwell time algorithms

The goal is to quantify the time spent in a shopping centre and a shopping area reliably and accurately.

To calculate Dwell Time, we process the same data that is used to quantify the number of visits. For each person entering and leaving the polygon, we calculate the time spent in the shopping centre or the commercial area by using the first identified timestamp and last identified timestamp of the person device in the polygon and then aggregate the data of individuals to showcase statistics on the platform.

As for the dynamic footfall analysis, visits are aggregated by day. So a person who comes on two different days is counted as a new visit.

On a given day and for a given polygon, the visit duration of an AID device is summed up. For example, a device that is present between 9:00 am and 9:30 am and then again from 11:00 am to 12:00 am in a particular polygon will have a total visit duration of 1.5 hours.

For calibration and validation purposes, we ran multiple sources to calibrate and validate this new metric:

• multiple coherence checks across assets and geographical areas
• and surveys across Europe to calculate theoretical Dwell Time

To improve the reliability of the data, for this analysis we have considered the opening hours to be 8 am-11 pm. Any visits starting before 8 am or ending after 11 pm are automatically filtered out and therefore excluded from this analysis.

This allows us to exclude visits not related to the commercial aspect of the centre (workers, cleaning crew,...).

Moreover, to guarantee the quality of our metrics, we filter out short visits that are harder to quantify and less considered engaged visits. So the data available on the platform only concerns visitors who stayed more than 15 minutes and less than 5 hours.

Mytraffic has historical data since August 2021: we began computing the metric in August 2021. If the polygon is more recent, data will be available from the moment the polygon is created, in the same way as for dynamic visitor analysis.