Customer base analysis

Goal: customise the homepage and the newsletters shown to each user in order to increase their propensity to purchase
Use case

Italian e-commerce

Implemented algorithms

Product categories

Clustering algorithm in relation to activity on the website

Temporal activity

Clustering algorithm in relation to activity on the website

Conversion

"Survival curve" algorithm for conversion/channel analysis

Data

The starting data are several years of user navigation and activity logs on the site, particularly relating to user interests and the temporal dimension of activities.

  • Creation of APIs that can be called from external systems that return results related to the user under consideration
  • Dynamic graphic interface for data exploration.