Personal Financial Manager

High conversion of banking product offers to bank customers implies a deep understanding and analysis of the behaviour of bank customers, while information on customer spending on bank cards is a valuable source of user data. Thematic modeling and segmentation of customers is carried out by decomposing the initial customer transactions into interpretable behaviour profiles, followed by highlighting the most attractive customer groups for the subsequent sale of products.

Input Data

  • Transactions of bank customers
  • Description of MCC codes and merchants
  • User accompanying information
  • Results of sales of banking products

Output Data

  • Bank User Behaviour Model
  • Prediction model of the probability of additional purchase of a banking product
  • Customer base segments

Main Results

A pilot project was carried out, integrated into the customer’s business process:

  • Identified subjects 80% of users of bank cards
  • 85% of consumption profiles are interpretable
  • Prediction of socio-demographic parameters of the audience
  • A two-level hierarchy of customer consumption profiles has been built
  • Solved the problem of finding similar consumers “Look-a-like”

Customer consumption clusters

Customer consumption clusters
Degree of coverage of interpretable types 80 %
Average number of consumption types 5
The proportion of interpreted consumption types 84 %
Average number of categories expenses in type 6
Average share of customers per type 0.19
Sample type “Sick”
Therapeutic Services 0.28
Licensed doctors 0.23
Dentists, orthodontists 0.09
Hospitals 0.02