Companies spend significant sums to acquire customers. Once hooked, marketers protect those investments by attempting to keep patrons happy, engaged, and most of all, loyal.
Reducing customer attrition, or "churn" in marketing parlance, often involves offering incentives such as discounts to individuals identified as likely to defect. The tricky part comes in figuring out exactly who should be targeted.
“You have to look at the net profitability of the retention campaign”
Sunil Gupta, the Edward W. Carter Professor of Business Administration at Harvard Business School, argues that companies often fail to take into account the complete value of the customers they are trying to retain.
"What's missing from traditional methods is that they focus only on a customer's likelihood to churn, but not on the overall profitability of that customer," says Gupta, who cowrote the working paper Managing Churn to Maximize Profits with Aurélie Lemmens, an associate professor at the Tilburg School of Economics and Management.
They propose a new method to target customers that potentially boosts a company's profits in a big way—by more than 100 percent on average when compared to standard retention targeting efforts.
The First Cut
Customer attrition is a widespread problem that affects firms in a variety of industries. For example, US credit card providers often deal with annual churn rates of about 20 percent, and mobile phone carriers in Europe battle 20 to 38 percent churn, according to the paper.
Lost customers lead to untapped dollars. A McKinsey report estimated that reducing churn could increase earnings of a typical US wireless carrier by as much as 9.9 percent. It's no surprise then that executives in both the United States and Europe say customer retention is their highest marketing priority—and they've been given bigger budgets to fight the battle.But as Gupta points out, identifying potential defectors to target is not as easy as it sounds. For years, companies have used churn management strategies that generally follow two steps: ranking customers based on the estimated likelihood that they will flee and then offering incentives to a core group of customers at the top of the churn ranking.
Gupta says these traditional methods are flawed. The two steps are not set up to maximize bottom-line profit of the retention campaign because they ignore a simple but important fact: Not all customers are equally important to the firm.
"You have to look at the net profitability of the retention campaign," Gupta says. "If I offer an incentive to customers most likely to churn, they may not leave the company, but will it be profitable for me? The traditional method is focused on reducing churn, but we contend the goal should be maximizing profits, rather than only reducing churn. People have been trying to refine and improve the method for the last 10 to 15 years, but many are missing the bigger picture."
To gain maximum profit from retaining customers, companies should consider not only the churn probability of customers, but also how much they spend, the likelihood that they will respond to a retention offer, and the cost of the offer itself.
Gupta's predictive model takes all these factors into account and also determines the optimal number of customers to target, rather than targeting an arbitrary number. Devising this number means finding just the right balance in the inevitable compromise the firm has to make between increasing the target size to reduce the potential loss of money from defection and reducing the target size to trim the cost of the incentives themselves. After all, while companies don't want to lose customers, targeting too many of them could become too expensive to be worthwhile.
“Acquiring a customer is far more costly than keeping a customer”
It makes sense to target higher-value customers who are most likely to defect and who are also most likely to respond to an incentive offer.
The model also takes into account customers' likelihood to respond to incentives. If a high-value churner is not likely to respond to a promotion, there's no need to send that person the promotion, Gupta reasons. "Sometimes the cost is low, like in the case of email. But sometimes it's substantial, where companies are sending high-gloss brochures."
Bottom Line
The new method certainly isn't perfect; Gupta says he actually saw more prediction errors in terms of which customers will churn. However, the new method leads to better predictions where it matters most for a company's profit by allowing companies to target their most valuable customers while factoring in the cost of the incentives.
"Even if we are a little wrong in predicting the likelihood of customers] to churn in some cases, it's OK. Our goal should not be to minimize the accuracy of churn prediction, regardless of who the customer is. Our goal should be to minimize the error in profitability of who we target," Gupta says. "Even if I'm worse in my churn prediction, I will still be better off."
During their research, Gupta and Lemmens applied this method of more precise customer targeting to a major US wireless carrier—a company not identified in the study. They found that the approach led to, on average, a 115 percent improvement in profit compared to traditional targeting methods. These extra profits came with no additional implementation costs for companies.
"All we did was use the same incentive you would otherwise use," Gupta says. "There was no change in anything except who you target, so there was no additional cost."
Gupta and Lemmens also used projections to show that Verizon Wireless, the largest wireless provider in the United States with 111.3 million subscribers, could see a profit increase of at least $28 million from a single retention campaign if it applied their approach.
Gupta says the method could be applied in nonprofit areas to predict, for example, patient compliance to a medical treatment. In fact, this gain-and-loss-based method should help just about any organization looking to retain customers with discount offers.
"I think anywhere where churn is a big deal and companies are spending resources to reduce churn, this will be useful," Gupta says. "The larger the customer base, the more beneficial this model will be to the company. From my view, acquiring a customer is far more costly than keeping a customer. So any company that wants to retain its customers should find some value in this."
Rather than trying to find better analytics and predictors of churn, shouldn't we invest our resources in being better companies? Providing for better customer experience?
I learned long ago that you will never maintain a long term competitive advantage based on price. You call it "incentive." It's what you do BEYOND the product or service that makes customers loyal. The reality is they can buy what you're selling from a myriad of places. It's what you're not selling but what you do for them that will reduce your churn.
Happy customers don't leave. Why would they?
But they miss a critical strategic issue: the value of a customer to a company goes far beyond the revenue that that that customer brings to the company.
Total relationship value must take into account a variety of metrics ranging from strategic or competitive value to disruptive potential.
But total relationship value is only half the picture. You also need to know likelihood of churn.
An earlier comment noted that "happy customers don't leave" - this could not be further from the truth.
The evidence is that happiness is only one factor in a customer's decision to stay with your company. Many happy customers leave. And many unhappy customers stay (think About why you keep flying your preferred airline or using your cable company - probably not because you think they are amazing)
Understanding why your customers make the choice for your service and why they choose you is critical. Christensen famously gave us a starting point for this in his assertion that everyone buys a product to do a job. What is your product's job?
I am fortunate to work with companies that understand that customer retention and profitability is about more than lifetime revenue value and happiness. And they're very successful.
Occasionally some customers may make demands which cannot be met. No need spending time, energy and resources to follow them too long.