3 ways predictive analytics can help financial services companies lower customer attrition

Upon arrival at work on a Monday morning, Joana, a relationship manager at a leading retail bank has her weekly meeting with her boss to discuss her targets. Last week, she noticed in her CRM that revenue from her existing customers had dropped down by 3% due to lost customers and decrease of business with the bank. Despite her best efforts to understand the causes, most surveys she sent remained unanswered and the handful of clients she called either did not answer or gave short and elusive responses. Not only there was not a clear path of action, but for many clients, it was already too late. They had already left for the competitors.

The issue with data-blind customer relationships management

According to Accenture, 18% of retail bank customers switched providers completely and 27% added new providers in the last five years. At the same time, relationship managers rely only on tedious, unreliable, not-always-answered customer surveys and if more, simple two-by-two charts are used, making relationship managers blind when dealing with customer attrition. Data-driven companies, on the other hand, take a multidimensional approach combined with predictive analytics on massive amounts of data to identify well in advance customers about to churn.

3 ways predictive analytics can help financial services companies lower customer attrition:

1. Detection of at-risk-of-leaving customers. Today, most relationship managers discover they have lost an account only after the former has already made the irrevocable parting decision. It is therefore too late to reverse the situation. Predictive models using massive historical data are game-changing. They are built upon a full view of the customer that should first be built by aggregating disparate data consisting of, for instance, customer profiles, credit scores, account balances, transactions, or loans take-out and tenure. It may also include data from call centers or personal online account logs. Afterward, machine learning algorithms are able to discover patterns, based on historical data and events and are able to live-assign individual customer attrition risk scores, and display them on business intelligence or CRM tools, allowing customer relationship managers to identify about-to-leave customers.
2. Identification of customer attrition root causes. Knowing only the at-risk-of-leaving customers is not enough information if relationship managers want to reverse the situation. By leveraging the full view of customers, clustering algorithms are able to group clients presenting the same at-risk-of-leaving pattern. This will help to focus only on the most important groups of customers and leads to a deeper understanding of the main behavioral dynamics and causes behind attrition.
3. Tailored recommendations to prevent at-risk customers from leaving. Once the attrition model has well identified the root causes for attrition, the system is also able to provide recommendations for immediate actions to be taken to reverse the situation, based on successful techniques with similar clients in the past. Uplift modeling detects populations of customers sensitive to a recommendation. For targeted operations, it helps identify groups of people who are likely to respond positively or negatively to a solicitation and targets customers accordingly.

The tale of the data-driven customer relationships manager

A few months later, Joana’s boss took the step to integrate predictive analytics into their CRM to help manage customer attrition. Since then, Joana receives a weekly report identifying customers with a high risk of attrition, helping her to better focus her commercial team’s efforts. Joana now knows in what direction to investigate to suggest to her Boss improvements on the bank processes and services, as the new system is also able to highlight possible root-causes for customer attrition. Joana knows from her experience that it takes time to completely transfer cash or assets and close a bank account, and therefore a turnaround is still possible. If done quickly, she could ensure continuity with the at-risk customers. She shoots a note to her boss, and they schedule an all-hands meeting to decide on actions to take based on the recommendations given by their new data-driven CRM.

If, in your company, you do not leverage predictive analytics to prevent customer attrition, you should do it now!

Reducing attrition in financial services is more important than ever for a company to thrive, it is a matter of survival:

• There are no virgin consumers left for the taking: in developed countries, the financial services penetration rate is almost 100% among profitable consumers — in the sense that every consumer is already using all services that are essential.
• Competition is more aggressive than ever: fintech, big technology firms (e.g.: GAFA…) and traditional players are all fighting for the same customers, for whom the choice of service providers and the easiness of switching between them is ever increasing.

Yet, many banks and financial services companies have to take the steps to build a strong analytical foundation for success. Those who decide to move will be best positioned for success in the future.