Measurements for anonymization and pseudonymization

Jan Antala
Pygmalios Engineering
3 min readDec 5, 2019

At Pygmalios, we value the principles of Data Privacy. We perform anonymization and pseudonymization of the data related to natural persons, so it is not possible to identify visitors or employees on the premises.

We perform anonymization and pseudonymization of the data

To secure privacy and compliance with legal requirements, we took following measurements and technical restrictions for every module.

Traffic

Advanced footfall analytics based on the data from mobile phones and accurate video sensors. By combining these two sources you can get a very accurate picture about footfall trends, overal location potential and basic customer profile.

Footfall Analytics by Wifi

  • MAC address is hashed in probe requests, different salt is used for each client
  • we create anonymous wifi sessions with configurations, where we define the maximal length of visit. All other data is filtered, so it is not possible to misuse the system for evaluation of employee behavior
  • we provide only aggregated metric data
  • metric data is collected only during opening hours, so we can exclude the possibility of evaluation of employee behavior

Footfall Analytics by People Counting

  • we collect only aggregated in/out data
  • we provide additional measurement for employee protection through tags, they ensure the employee is not monitored. Even without using the tags, employees are impossible to be identified
  • metric data is collected only during opening hours, so we can exclude the possibility of evaluation of employee behavior

Browsing

Detailed movement and zone analytics provide shopping area and product zone visit insights. Customer journey is analysed using Real Time Location System by cameras or bluetooth tags.

Movement Analytics by Real Time Location System

  • we collect coordinates only in defined shopping areas with exclusion of backoffice and cash desk
  • metrics are analysed from anonymous browsing sessions (full customer journey) defined with special initial and end point zone (typically the entrance and exit). We define the maximal length of visit, other data is filtered, so it is not possible to misuse the system for evaluation of employee behavior. For the same reason, we filter also the backoffice area.
  • we provide additional possibility to set up exclusion of critical zones in shopping areas
  • metric data is aggregated and collected only during opening hours, so we can exclude the possibility of evaluation of employee behavior

Demographic

Adjust your marketing activities to your real target audience. Thanks to the dedicated video analytics sensors, you will be able to analyse the gender and age structure of your customers with accuracy ±5 years.

Gender and Age Analytics

  • we do not store any video or image, we only collect metadata about visitors necessary for identification of session. Only metadata is being processed, so we can exclude the possibility of identification.
  • we use only aggredated data of approximate age and gender
  • we monitor only the entrance to the premises, not the shopping areas inside the shop
  • metric data is collected only during opening hours, so we can exclude the possibility of evaluation of employee behavior

Queue Monitoring

Never ending queues lead directly to reduced shopper satisfaction. This results in low customer loyalty. Active queue monitoring mitigates these risks, reduces your customers’ frustrations and helps you better organise your staff. Accurate video sensors are used to monitor congestion patterns and enable you to forecast when and where queues occur.

Active Queue Monitoring

  • we collect only aggregated data about number of people and visit duration in the selected zones
  • we secure the exclusion of employees by defining the special queue monitoring zone
  • metric data is collected only during opening hours, so we can exclude the possibility of evaluation of employee behavior

Buying

Add your core productivity KPIs as a layer on top of the core behavioural KPIs and see how they relate to each other. Integrated with a large variety of POS systems, this module provides you with a detailed view of the essential business metrics (Revenue, ATV, UPT). Bundled with the Traffic module, it also allows you to monitor the critical indicators of store productivity such as the conversion rate of visitors into customers.

POS Data Integration

  • only aggregated data is available
  • metric data is collected only during opening hours, so we can exclude the possibility of evaluation of employee behavior

--

--