Case Study of a Financial Institution-Money Exchange and Remittance.

Advanced data analytics and econometric techniques were used to develop a Predictive Machine Learning Model by identifying Customer Loyalty Indicators based on a large historical database of customer profile and their remittance transactions over last 2 years. The client-a leading money exchange and remittance organization was looking to reduced customer churn in todays highly competitive market and wide choice of remittance methods.

Over 15 variables like customer demographics, remittance transaction frequency, amounts, locations and many more were mapped, measured and analyzed to identify the best predictive indicators for Customer Loyalty .

We developed a Sophisticated Predictive Model based on eight feature rich variables to forecast potential customer churn which could identify the customers more likely to leave in the next 3 months. The model also categorized the customers into different segments based on their expected loyalty level mapped to their top churn reasons.

Using the segmentation model, the client is able to identify the customers based on their loyalty levels and is able to offer them special promotions corresponding to their churn reasons. These campaigns have  resulted in an increase in customer loyalty while reduction is noticed in customer churn levels over the last six months. It has also helped to improved the effectiveness of their Marketing Campaigns in terms of increased returns and uptake.

The model is being continually monitored and enhanced using machine learning algorithms. Efforts are being made to identify more Loyalty Indicator Variables and collect robust data for them.

The enhanced model is expected to reduce customer churn significantly and also result in reduced cost of promotions and new customer acquisition and hence increase the Marketing ROI.