Attracting and Keeping the Customers You Want with Predictive Analytics

    

How do you attract the customers you want? How do you keep them?

Let’s consider the case of an on-line bank trying to attract new deposit customers—particularly those with larger balances available to invest. The bank decides that it’s willing to take a hit in margin to grow the volume of balances and institutes a promotional rate (over and above the underlying rate) that will last for 12 months. After 12 months, the customer loses the bonus and their rate reverts to the underlying rate. While the bank had been employing this strategy for some time, it wants to know whether a substantial increase in the bonus rate would attract more and different customers.

The bank instituted the test and was pleased with the response (see the graph below). A dramatic increase in acquisitions was the result, and it was not only additional customers, but also customers with a higher average balance. So far, so good.

pic 1 for post
But what happened when the bonus expired? The graph below shows attrition rates, leading up to bonus expiry and for the few months after. The pink line represents the typical pattern the bank had experienced with its bonus programs in the past. The turquoise line represents the current test. While the dramatic bump in rate attracted those customers who were most price sensitive, twelve months later, those customers were still the most price sensitive and were most likely to leave as soon as the bonus expired.

pic 2 for post

The role of predictive analytics

Big data and predictive analytics are big topics of discussion. Here’s an example of where they can really make a difference. In this case, predictive analytics allows you to quantify the expected movements of balances in a way that eliminates some of your uncertainty. Being able to predict which customers are more price-sensitive to acquisition rates, as well as with respect to attrition, lets you plan for trade-offs and ultimately make better long-term decisions.

For example, the bank might choose to aggressively attract price-sensitive customers, and then provide an incentive them to move to a stickier product (such as a fixed-term deposit) as their bonus expires.

Or the bank might choose to balance the lost margin from the more price-sensitive customers, by simultaneously reducing rates for customers who care about factors other than price (certainty of returns, better service, etc.)

Knowing (vs. guessing) allows you to make the optimal tradeoffs. While nothing is certain, the increased information can allow you to navigate the choppy waters ahead with increased confidence.

About The Author

Seasoned executive with breadth of experience in international business development, P&L management, operations, strategic planning, and R&D. Broad and deep expertise in developing underlying infrastructure for analytically-based decision-making; recruitment, development and retention of analytic talent; development and implementation of methodology and processes to efficiently develop internal predictive and decision models; and successful management of analytic research and development from selecting the right projects through transformation into successful products. Specialties:research and development, product management, data mining, predictive modeling, decision modeling, optimization, credit risk management, customer value optimization, pricing, analytic software, solution design, payment products, fraud detection and prevention, analytic technology