|
By Lisa Yorgey Lester
How to identify possible churn within your database.
Customer defection should come as no surprise. Your customers are waving goodbye as they slowly walk out the door. If you learn to read the signs beforehand, you can launch a preemptive strike to potentially keep them from churning—before you need to spend additional money to win back customers in whom you've already invested. That is, if they are profitable. If not, you may simply want to let them keep walking.
Suss Out Churn
At its essence, segmentation divides a file into groups that behave differently. In this case, you want to find the segment of your customers that is likely to churn, and then develop a program that leads to measurable increases in retention and profitability.
When looking at churn, you want to divide your customers into four quadrants: those who have high lifetime value that are likely to churn or unlikely to churn, and those customers who are less profitable and likely to churn versus those unlikely to churn.
To determine which group of customers is likely to defect in the near future, you need to identify the signs that predict churn. Start by closely examining a sample of customers who recently have defected. "Look at the characteristics of who they were when they left, and what was happening prior to them leaving," explains Don Hinman, executive vice president, product development, and senior principal of professional services with Allant, a marketing optimization solutions provider based in Naperville, Ill. What makes this group homogenous? A pattern of decreasing revenue or number of products purchased often are signals that a customer is getting ready to exit your business. A financial institution, for example, might see that customers who defected had decreasing account balances or stopped using their credit card in the six months prior to closing their account. If you add demographic or lifestyle data, you might also find these defectors tend to be younger rather than older, Hinman explains.
According to Sam Koslowsky, vice president of modeling solutions with Harte-Hanks, a direct and targeted marketing company based in San Antonio, there typically are five to 10 data elements combined together that provide a model, which can predict who's likely to attrite. While the precise mix of elements often depends on your vertical market, they may include changes in account balances, the number of transactions from year to year or purchasing behavior. You also may want to consider the tenure of customers who've recently defected. Are these new or established customers? Or, in the case of a retailer, have the types of products purchased changed? Do these defectors buy the same thing from year to year, or are they buying new things?
How to identify possible churn within your database.
Customer defection should come as no surprise. Your customers are waving goodbye as they slowly walk out the door. If you learn to read the signs beforehand, you can launch a preemptive strike to potentially keep them from churning—before you need to spend additional money to win back customers in whom you've already invested. That is, if they are profitable. If not, you may simply want to let them keep walking.
Suss Out Churn
At its essence, segmentation divides a file into groups that behave differently. In this case, you want to find the segment of your customers that is likely to churn, and then develop a program that leads to measurable increases in retention and profitability.
When looking at churn, you want to divide your customers into four quadrants: those who have high lifetime value that are likely to churn or unlikely to churn, and those customers who are less profitable and likely to churn versus those unlikely to churn.
To determine which group of customers is likely to defect in the near future, you need to identify the signs that predict churn. Start by closely examining a sample of customers who recently have defected. "Look at the characteristics of who they were when they left, and what was happening prior to them leaving," explains Don Hinman, executive vice president, product development, and senior principal of professional services with Allant, a marketing optimization solutions provider based in Naperville, Ill. What makes this group homogenous? A pattern of decreasing revenue or number of products purchased often are signals that a customer is getting ready to exit your business. A financial institution, for example, might see that customers who defected had decreasing account balances or stopped using their credit card in the six months prior to closing their account. If you add demographic or lifestyle data, you might also find these defectors tend to be younger rather than older, Hinman explains.
According to Sam Koslowsky, vice president of modeling solutions with Harte-Hanks, a direct and targeted marketing company based in San Antonio, there typically are five to 10 data elements combined together that provide a model, which can predict who's likely to attrite. While the precise mix of elements often depends on your vertical market, they may include changes in account balances, the number of transactions from year to year or purchasing behavior. You also may want to consider the tenure of customers who've recently defected. Are these new or established customers? Or, in the case of a retailer, have the types of products purchased changed? Do these defectors buy the same thing from year to year, or are they buying new things?
Companies Mentioned:
 |
|
|
