Emerge's churn models form part of retaining existing clients.
In the banking industry, by the time the bank is notified that one of its clients is leaving it is usually too late to save the client – he or she has already opened another account elsewhere and moved funds and debit orders to the new account. No amount of persuasion will help at that point. Being able to identify dissatisfied clients some time before they leave is essential to be able to deploy a client-retention strategy effectively and cheaply.
Emerge Analytics trained a machine learning model for a South African bank that was losing in excess of 8,000 current account holders per month. Implementing a client-retention strategy on its entire base of several million account holders was simply not feasible. However, Emerge Analytics’ model allowed the bank to identify more than 50% of the account holders who would leave in the next three months by selecting a group of 0.5% of the client base. These results are not just theoretical but are based on real life experience.
The saving of both interest- and non-interest revenue to the bank amounted to several million Rands each year.