BANKING CHURN PREDICTIONCUSTOMER CHURN ANALYTICS TO MANAGE RISK IN BANKING
Tackling customer churn to identify and manage risk in the banking sector
Churn analysis identifies $67mn of ‘at risk’ customer business
A bank struggled with customer churn which was higher than the industry average. Predicting which clients will churn should form a crucial part of any organisation’s client-oriented strategy, as the cost of acquiring new clients is far greater than the cost of keeping existing clients.
Retaining consumers is easier when an organisation understands its’ clients in a holistic way. Early warning signals that might point to attrition can be the cancellation of automatic payments, client complaints on client calls or comments on social media.
By analysing the different indicators, we enabled our Client to identify and address potential client churn before it was too late
Deploying predictive analytics to calculate churn propensity
The Client had a hindered view of customer sentiment and needed to better understand what and why customers were at risk to take approrpiate forward action.
A robust analytics approach was required that could scrutinise multiple data sets and establish a commonly accepted definition of customer churn, to support performance validation and impact assessment.
To find a robust method by which to compute Customer Churn propensity and to map out Churn timing.
To develop a Decision-support system that would incorporate multiple attributes including, customer value, profitability and customer profile.
We needed to idenitfy early warning signals that would point to attrition include auto. payments cancellation, client complaints or comments on social media.
By analysing the different indicators, potential client churn could be identified and addressed before it was too late.
Our solution targeted the dynamic generation of customer churn scores on a probability and risk rating basis.
Our algorithms could handle a wide array of fields and data types, allowing clear identification of the fields most important when predicting churn.
This yielded invaluable information to determine not only which clients were churning but why and when they were anticipated to leave.
The Bank could then accurately identify over 12% of accounts within the selected portfolio of branches that would likely close within the following 3 months and take according steps to prevent them from churning.
Risk identified for forward action by client
Absolute reduction in Customer Churn