Predicting a Cure for Collections-Induced Customer Attrition

Why are we writing an article about customer attrition for collections professionals? After all, customer attrition and retention are the tasks of the customer service and marketing departments, right? No, not anymore. Today customer retention is a critical issue for every customer-facing department in companies that operate in highly competitive industries, such as telecom, banking, insurance, and health care — where customer attrition is climbing at an accelerating rate.

In this increasingly competitive environment, customers are switching from one company to another with surprisingly little motivation — a slightly lower interest rate, a free bonus offer, or a single rude customer service call. So, imagine the negative impact that a collections call can have on some customers.

Customer retention may seem inconsequential compared to collections’ primary objective: Collecting more outstanding debts. After all, delinquent accounts are an enormous source of revenue loss, so you must take every action necessary to reduce losses and boost revenue. That’s true, but long-term revenue is also lost when potentially good customers are so annoyed that they sever their relationships with your company. We’ll call this phenomenon "collections-induced customer attrition."

But now collections organizations can both improve their collections results and help support customer attrition using predictive analytics to determine who to call and who not to call.

Who Are You Going to Call?
No doubt you currently rely on behavior scores to rank your collection accounts according to risk. This traditional analysis helps organize your daily activities so that you can target your resources to reach riskier accounts sooner. However, behavior scores can only give you a one-dimensional customer segmentation capability. In other words, they can only lump accounts into broad categories where, for example, you decide that agents call accounts within a particular risk range and letters are sent to accounts in another risk range.

Behavior scores cannot give you, however, the ability to see each individual account’s propensity to react to the collections action you take. This is a big loss, because predicting customers’ reactions is the key to identifying accounts that are likely to self-cure. And identifying self-curing accounts is the key to achieving several strategic benefits including:

• Decreasing your agents’ workload
• Increasing collections results
• Reducing collections-induced customer attrition

Most collections departments have implemented many ways to achieve the first two of these three benefits, but not the third. However, a growing number of collections organizations are realizing that they can, and should, contribute to companies’ efforts to retain good customers. They are achieving this goal by targeting customers who are likely to self-cure in the early-stage of collections, then removing them from their call lists.

Predicting Who to Call — and Not Call
To more fully understand the advantage of identifying self-curing accounts, let’s look at three general types of customers who land in collections:

• Joe habitually spends more money than he earns. As a result, he has many outstanding bills each month and only a limited income with which to pay them. His payment strategy is typically to quiet the bill collectors that reach him first.
• Tom, like Joe, spends more than he makes. But unlike Joe, Tom doesn’t care who calls or how often: He’s not going to pay anyone. His philosophy: You can’t get blood from a stone.
• Sam is nothing like Joe or Tom. His expenses pretty much match his income. But for one reason or another he’s occasionally late paying some bills. He views himself as a good customer, not a deadbeat. And if you call him about a late payment and make him feel like a loser, guess what? The next time he gets another offer from one of your competitors in the mail, over the Internet, or by telephone, he just might jump at the offer.

This simple overview of the personalities and attitudes of three basic types of collection accounts reveals a very important clue to supporting customer retention: Sam is highly likely to self-cure, so do not call Sam. Instead, give him a chance to pay on his own. In this way, you’ll reduce the chance of insulting, humiliating, or angering him — and all of the other Sam’s in early-stage collections.

From this example, it’s easy to see that when you gain the ability to predict your customers’ reactions to your collections actions, you gain the power to not only make the most of your resources, but also to make a significant contribution to your company’s overall customer retention efforts.

What’s more, not only will you help prevent customers from closing your accounts, but also any other accounts they have with your company. For example, banking customers may have checking accounts, credit cards, auto loans, and mortgages with one bank. If they are unsatisfied with one department’s actions, the sentiment will carry over to the company as a whole; making them more likely to take their business elsewhere.

Gaining the Advantages of Self-Cure Prediction
One of the nation’s leading MasterCard and Visa issuers realized the value of predicting its collections accounts reactions to its collections actions. The collections department deployed action-specific predictive analytic models to compare the probability that each account would cure with a call against the probability that the account would self-cure without a call. With this insight, the company removed the customers identified as self-cures from its early-stage call list.

The company immediately reported significant improvements in these critical areas:

• Each month balances cured and dollars collected increased
• Agents’ time was focused where their efforts would have the most impact
• Collections avoided offending customers who were likely to self-cure

These benefits will translate into any collections organization that identifies early-stage accounts with the propensity to self-cure. Without deploying this capability your company will continue to pay a high price for taking the risk of offending these customers.

Clearly it’s time to give the Sam’s of the world the opportunity to self-cure — and time to lower your collections department’s risk of causing collections-induced customer attrition.

Robert Tate is the vice president of marketing at Austin Logistics (www.austinlogistics.com). His responsibilities include corporate communications, product positioning and demand generation activities. Tate holds an undergraduate degree from the University of Southern California and a graduate degree from Duke. Bob can be reached at btate@austinlogistics.com.

Austin Logistics is a leading provider of analytic software and solutions that predict customer value and behavior to drive more profit from every customer interaction — delivering a proven return on investment in six months or less.