Leveraged Predictive Analytics to save wholesale bundle purchase cost

Customer:Leading European MVNO

Scope: MNO wholesale costs typically covers up to 40-50% of the total cost structure of a MVNO directly impacting profit margins. Sasken is involved in development and management of the analytics platform to predict usage at a subscriber level to achieve wholesale bundle optimization.


Prediction of usage at subscriber level is done with the following attributes:

  • Varied Data Sources: Aggregation of high volume of CDR data, subscriber data & product data available in discrete sources
  • Statistical Model with very high accuracy & usage prediction at a subscriber level: Usage prediction monthly, prescribe wholesale bundle to be purchased from MNO and self-correct as the model matures
  • Automated execution of statistical models and report generation
  • Accurate Reports: The reports can be leveraged for refining models and billing the client


  • Solution deployed in multiple countries in Europe such as Germany, Spain and UK. Solution and is running over the last 2 years
  • The solution saved the client over € 700,000 per month which is around 4% of the client’s revenue

Accuracy of Sasken models was very close to the perfect foresight (PF), i.e. reactive estimates of cost after the usage cycle is complete

Sasken Solution

Customer:UK MVNO

Customer:Leading American Industrial OEM