Do You Know Who Your Most Profitable Customers Are?

Data mining is helping companies pinpoint them

Like most companies, Federal Express Corp. has some customers that are losers: The cost of doing business with them is greater than the profits they return to the shipper. But FedEx has an edge that most companies lack--it knows who those customers are. That knowledge has kicked off a marketing revolution inside FedEx, where customers sometimes are rated as the good, the bad, and the ugly. "We want to keep the good, grow the bad, and the ugly we want nothing to do with," says Sharanjit Singh, managing director for marketing analysis at FedEx.

It's relatively easy for companies to recognize who is spending money with them. But surprisingly, identifying which customers are profitable for them requires a quantum leap in sophistication. Confounded by technical difficulties and the high cost of software and data storage, many businesses muddle along with only a rough approximation of which customers make them money. "Not too many people have actually done this yet because it's difficult to pull off," says Wayne Eckerson, a principal analyst at Data Warehouse Institute.

But those few that have begun are finding that determining each customer's profitability can pay big dividends. Savvy companies are conducting behind-the-scenes beauty contests to determine who they like and who they wish would go away. Companies such as U S West, Bank of America, and The Limited are creating vast information warehouses stocked with customer data that allow them to compare the complex mix of marketing and servicing costs that go into retaining each individual consumer, vs. the revenues he or she is likely to bring in. "Some companies are building a profit-and-loss statement for each customer," says Mike Caccavale, president of Aperio Inc., a consulting firm that specializes in customer-relationship management.

Of course, companies have been using large databases to help refine their marketing efforts for years. But previous attempts focused on the average behavior of large demographic groups, while the newer efforts allow marketers to target individual customers with pinpoint accuracy. And the latest techniques go well beyond determining whether a marketing campaign encourages a consumer to buy--they focus on whether or not that consumer spends enough to make a campaign worthwhile.

At FedEx, customers who spend a lot with little service and marketing investment get different treatment than, say, those who spend just as much but cost more to keep. The "good" can expect a phone call if their shipping volume falters, which can head off defections before they occur. As for the "bad" --those who spend but are expensive to the company--FedEx is turning them into profitable customers, in many cases, by charging higher shipping prices. And the "ugly"? Those customers who spend little and show few signs of spending more in the future? They can catch the TV ads. "We just don't market to them anymore," says Singh. "That automatically brings our costs down."

The power of such an approach lies in a company's ability to determine how much to spend on marketing campaigns--and where to direct those dollars. Twice a year, U S West sifts through its customer list looking for money-losers who have the potential to be more profitable in the future. The starting point is a database containing as many as 200 observations about each customer's calling patterns. By looking at demographic profiles, plus the mix of local vs. long-distance calls, or whether a customer has voice mail, they can estimate a customer's potential telecom spending.

Next, U S West determines how much of the customer's likely telecom budget is already coming its way. Armed with that knowledge, "we can set a cutoff point for how much to spend marketing to this customer before it begins to deteriorate our profitability," says Dennis J. DeGregor, chief database marketing executive for U S West. Those savings fall straight to the bottom line.

Seems like the kind of information marketers would die for. But so far, only a handful have taken the plunge. For one thing, it's expensive. Fleet Bank, one of the few big companies to break out its costs, is spending $38 million just to get started. Also, details about which customers are profitable and which aren't is not always welcome news. "It means the things you did in the past were wrong," says Scott Nelson, an analyst at Gartner Group. "Big advertising campaigns, product introductions--all might be mistakes because you find out they didn't contribute profit."

WRONG REVENUES. That's what executives at Glasgow-based Standard Life Assurance, Europe's largest mutual life-insurance company, discovered earlier this year. They were stunned when the first cut of a profitability survey showed that the insurer was loading up fast on policyholders who held little or no potential for making money. Instead of bringing in the affluent customers Standard Life wanted, its direct-mail marketing campaign was encouraging elderly couples and stay-at-home mothers to sign up for costly home visits by sales agents. Revenues were up--but they were the wrong kind of revenues. "It was an exact mismatch," says Graham Wilson, Standard's database and statistics manager. "These are people who love to sit down and have a cup of tea with someone, but they typically buy only one policy, and margins are small."

Even motivated companies may find mining for profitable customers a tough task. The mechanics are demanding. To get a true picture of customer profitability, Bank of America calculates its profits every month on each of its more than 75 million accounts. Mortgages, for instance, have different costs than checking accounts, car loans, or home-equity credit accounts. And service costs vary by how customers bank--whether they use ATMs, tellers, or the Net.

By wading through all that data, however, BofA is able to zero in on the 10% of households that are most profitable. It assigns a financial adviser to track about 300 accounts at a time. Their job: to answer questions, coordinate the bank's efforts to sell more services, and--perhaps most important--watch for warning flags that these lucrative customers may be moving their business elsewhere.

If the bank's computer notes a change in the customer's normal pattern of transactions or a falling balance, for example, it alerts the account manager, who may post a flag to the tellers at the customer's branch warning that the account is in danger of moving.

The next time the customer walks into the bank, a teller asks if they have any concerns about the account and offer them a chat with the bank manager. The heavy intervention seems to be working. Since BofA launched the program 18 months ago, customer defections are down, and account balances in the top 10% have grown measurably. "When such a small percentage of customers generate such a large percentage of profits, this kind of program is critical," says Christopher Kelly, senior vice-president for database marketing at Bank of America. "We view it as a strategic weapon."

HANDHOLDING. Often, the strategy is used not to jettison low-spending customers or to reward the cream of the crop. Efforts are focused on those in the middle--customers who aren't yet profitable but whose demographics suggest that they could be. Consider Joseph Taylor, a 62-year-old minister in the United Church of Christ and a professor of pastoral theology at Howard University in Washington. When First Union Bank bought Signet Bank in 1997, Taylor's accounts moved, too. As First Union combed through Signet's records looking for customers with higher profit potential, up popped Taylor. Soon, a marketing push to profitability began.

First Union's buyout of Signet was the third time Taylor's bank had been acquired, and he was fed up. But he dropped his plans to switch when he received a phone call from a First Union marketer. "He talked as if he had known me, even though we had never met," Taylor says. Taylor maintained high balances in his bank accounts, frequently writing large checks to help church members in need. The marketer suggested that Taylor consolidate his checking and savings accounts in one interest-bearing account, providing income to Taylor while reducing the bank's servicing costs. First Union also persuaded Taylor to sign up for a home-equity loan, and he brought over the accounts from his ministry as well.

Customers who cost more to lure than they're worth aren't likely to see the same pampering. For those who enjoy browsing through catalogs without buying much, the days of receiving free reading material may be numbered. Mail-order retailers such as Victoria's Secret are winnowing their lists to focus on the most profitable recipients. "Some we expect to drop--those we don't get any return from at all," says Frank Giannantonio, chief information officer for Victoria's Secret.

"Free money" offers are already on the wane. MCI Communications, Sprint, and AT&T have largely dropped their once-omnipresent offers of cash for customers who switch their long-distance service. "They were offering $20, $30, $50 to all takers, without any understanding about whether they actually would want to retain those customers once they had won them," says Thomas Tague, chief operations officer for Tessera Enterprise Systems Inc., a data-warehouse builder. Now, focusing on profitability gives the phrase "valued customer" a whole new twist.

Before it's here, it's on the Bloomberg Terminal.