RiskRobo

This services uses customized robots fully tailored to the objective

Our core competence on data and AI focuses on credit risk management combined with marketing management.

Case study

Challenge

A second tier bank in Europe has models in place to predict churn on a product level. However the marketing and service department struggle how to act upon the customers that are identified to be likely to churn. Based on their earlier experience with the CreditRobo of AdviceRobo they asked AdviceRobo to support with further developing the model whilst adding Psychographics. So, a better customer profile can be developed which makes it easier to act upon.

Approach

A sample set of customers (including churned customers) were invited to fill in a Psychographic questionnaire measuring not only ‘willingness to pay’ but also measuring other traits and behaviour on financial decisioning as well as satisfaction with the banks services.

The bank delivered the performance data of these customers and AdviceRobo combined these with the Psychograhic scores and profiles.

AdviceRobo applied a hybrid data modelling approach for both clustering (low, medium, high propensity) and classification (profiling for driving campaigns). In doing so AdviceRobo trains the models applying different machine learning modeling approaches to select the champion model.

Results

This case is in current development. Result will be available soon.

When the champion model is ready the next phase will be to design and apply treatments and automate these.