Customer Churn: A Key Performance Indicator for Banks:
“In 2012, 50% of customers, globally, either changed their banks or were planning to change. In US and Canada, customers who changed their banks increased from 38% in 2011 to 45% in 2012.” – Global Consumer Banking Survey 2012, Ernst & Young.
Customer churn and engagement has become one of the top issues for most banks. It costs significantly more to acquire new customers than retain existing ones, and it costs far more to re-acquire deflected customers. In fact, several empirical studies and models have proven that churn remains one of the biggest destructors of enterprise value.
A recent Forbes magazine article on lack of understanding of customers by CXO’s states — “C-level executives estimated that the lack of positive, consistent and brand-relevant customer experiences could cause them to lose out on a staggering 20% in annual revenue”. We are talking about hundreds of millions of dollars in revenues lost for any sizable banking enterprise.
Essentially, understanding your customers’ needs, preferences, sentiments, behavior and propensity to switch has become paramount for banks.
How SoLoMo is affecting customer sentiment and churn?
In today’s interconnected world the bad news spreads rapidly via exploding social media interactions. According to the Ernst and Young survey, in the US, 63% of customers use online personal networks and communities as a trusted source for information on banking products. And 45% of customers comment on the level of service they received in social media. The ability to track customer sentiment can give banks early indicators into customer service or pricing issues.
However, the information about customers’ sentiments and their experiences across multiple channels lies in many structured and unstructured data sources. More importantly the information is almost always locked within functional and application silos. This makes it challenging for the banks to get a holistic understanding of their customers, detect early warning signs and engage them with retention offers.
The key issue: knowing the customer and predicting churn:
In order to identify early signs of potential churn you first need to start getting a holistic 360-degree view of your customers and their interactions across multiple channels such as bank visits, calls to customer service departments, web-based transactions, mobile banking and social media interactions. This would allow you to detect early warning signs such as reduced transactions or stoppage of auto-pay or negative experiences, and you can take specific actions to prevent churn.
However, the totality of information about the customer is typically locked within functional and application silo’s that makes it very difficult to first detect early warning signs and then take actions to course correct, in real time. As a result, the banks end up strategizing and operating on basic slices of incomplete information from each individual silo. making themselves vulnerable to churn and massive revenue loss.
How Big Data Can Help you predict potential churn
The growth in volume, variety and velocity of data generated about the customers and their interactions across multiple channels has made it almost impossible to store, analyze and retrieve meaningful insights using traditional data management technologies.
But now, Big Data can help you solve these challenges and allows you to leverage both structured and unstructured data from multiple channels such as bank visits, customer call logs, web interactions, transactional data such as credit card histories, and social media interactions.
Native Big Data technologies solve the data management challenges by storing, analyzing and retrieving the massive volume and variety of structured and unstructured data economically on commodity hardware and scale elastically as the data grows. It also allows banks to tap in to real-time customer interactions that are more likely to provide early warning signs before it is too late. Additionally, sophisticated data matching capabilities allows banks to eliminate the data silos, connect the dots of a customer’s interactions across multiple channels and build a comprehensive holistic customer profile to gain a real-time, 360-degree view.
It all starts with knowing your customer holistically by unlocking the slices of information from multiple silos and turning them into actionable 360-degree customer intelligence.
Building a churn prediction model:
The big question is – would the 360-degree customer view alone be sufficient for a bank to predict potential churners in an efficient manner? By leveraging the holistic customer information, you need to build a viable churn prevention model. The churn prediction model with high quality score will arm you with the insights to identify the high-risk “real” churn targets and eliminate the “other” churners such as bad payers. Additionally, for each churn model, you can create lift charts to graphically represent the improvement that the model provides against random targeting of high-risk customers.
Being able to predict the churn and the right segment to target is not enough if you cannot determine marketing offers that should be delivered to each individual customer. Generic marketing offers based on broader segmentation lead to lower redemption rates. You need to be far more precise and make the offers very targeted and personalized to see a sizable reduction in churn. Machine-learning algorithms such as Collaborative Filtering can be very effective to offer very personalized marketing recommendations.
In a nutshell, customer intelligence management based on deep business process knowhow, and the use of Big Data and sophisticated machine learning give banks a distinct competitive advantage with an ability to predict and prevent churn, drive cross-sell and build customer loyalty. What are you doing about churn?