|
||
 |
||
|
When Dennis Liberson flew into Washington to interview for the top human resources position at Capital One Financial Corp. (www.capitalone.com), he was told that his interviews with the 16 top executives would have to wait. First, he was whisked off to a hotel room to take an algebra exam and write a business plan. Liberson, who jokes nearly seven years later that he might have been “the only HR guy who could pass their math test,” got the job and is now one of the company’s executive vice presidents. He also got an early taste of Capital One’s obsession with testing. The company, which pioneered such widely copied ideas as lower introductory interest rates for credit cards, tried out 45,000 different credit-card offers in 2000 to figure out how different groups of customers respond to new services and features. Capital One tests to find out which types of customers will respond to which offer. It tests which are most likely to use their cards, which pay their bills, and which pay on time. Capital One tests job applicants at every level to determine whether they are likely to make a good fit with the company. It even tests the best way (politely or otherwise) to collect money from customers who fall behind on their payments. “The number of things we’ve learned from testing is countless,” Capital One President Nigel Morris says. As for giving executives math tests, Morris says: “We found a lot of executives would be very good thinking on their feet, then you’d get analytical and they’d get vertigo.” Morris says testing is a cheap way of finding out customer preferences before rolling out a new offering or product—and he has the financial results to back him up. Capital One has reported 17 consecutive quarters of record earnings and has posted earnings increases of more than 20% in each of the past six years. Net income rose 30% to $470 million in 2000. The amount Capital One had loaned out to cardholders grew to $38.5 billion in the third quarter of 2001, up from $29.5 billion at the beginning of the year. Capital One, though still a young company, is “one of the best companies in that business,” says Robert Napoli, an analyst at ABN Amro, a financial-services firm (www.abnamro.com). “They probably use information and data better than any other company.” Traditional banks have tried to copy Capital One’s strategy of testing but they haven’t been as successful because “it’s not their total focus and passion,” explains Michael Auriemma, president of a credit-card consulting firm that bears his name. “They [Capital One] really believe in what they’re doing, and it’s ingrained in their culture.” Morris and Capital One Chairman Richard Fairbank hatched the idea for the company in the late 1980s while working as strategy consultants. After looking at a client bank’s credit-card operations, they figured they could do better. Banks were peddling one-size-fits-all credit cards, with the same 19.8% interest rate and a set annual fee. Banks, making good money, scoffed at the notion that they needed to change. Still, Fairbank and Morris thought they could use technology to segment customers and tailor different offers for different groups. They figured they could skim off the best customers from each group by giving them a better deal than they got from credit-card companies taking a cookie-cutter approach. At that time, the kind of database technology that could enable their strategy was just becoming available from companies such as Oracle Corp. (www.oracle.com). The credit-card field, moreover, was wide open. Banks had shunned nearly half the country’s population, denying plastic to those with blemished credit records or no credit records at all. “The more people who said it couldn’t work, the more animated we became, because we knew it would be too difficult for them to do it once we had,” Morris says. He says he and Fairbank decided they couldn’t just bolt their idea onto the side of a bank because the kind of unusual, information-based strategy they envisioned “has to be at the core of who you are.” They crisscrossed the country trying to find a backer. Finally, Signet Bank, a regional bank in Virginia, signed on. Capital One became independent through an initial public offering in 1994 and then a spinoff of remaining shares to Signet stockholders in 1995. To find and attract the most profitable customers, Morris and Fairbank pursued a variant of a strategy that has long proved successful for airlines and hotels and that is now spreading into telecommunications and many other industries. [See “Getting to Know You,” Context, December 2001/January 2002.] For airlines and hotels, which call the strategy “yield management,” companies use tests to tailor their prices to get the maximum revenue out of, in particular, business travelers. In telecommunications, where the strategy is sometimes called “customer value management,” service providers test to see which incentives, such as free calls in off-peak hours, will do the most to boost revenue. At Capital One, executives take a highly scientific approach to their testing. “We don’t just throw 45,000 random spaghetti at the wall to see what sticks,” Morris says. Instead, it slices up the population into tiny segments and pitches slightly different products to each, changing its approach as it learns what works and what doesn’t. Capital One tries out different interest rates. It tests to see what happens if it tells customers the rates will rise after various numbers of months, or stay fixed. It offers varying annual fees. It offers a slew of special-interest cards, such as for chess players or those of Irish heritage, whose cards carry a picture of the Blarney stone. Capital One even tests what happens if it moves a key sentence on the back of the envelope for a direct-mail piece. For customers with poor credit histories, Capital One says it has been cautious. It tests cards that carry minuscule credit lines or that sometimes require payment before the card can be used. Rivals, such as Providian Corp. (www.providian.com), are reporting a scary rate of default by so-called subprime borrowers because of the recession, so Capital One may still face problems. But ABN Amro’s Napoli says that “all signs are they are managing their credit quality extremely well.” Based on the latest figures available, Capital One’s loan losses amount to only 3.9% of its loan portfolio, far below the industry average of about 6.5%. Armed with the results of its tests, Capital One brings new variants to market so frequently that 95% of the products it offers today didn’t exist two years ago; 80% didn’t exist a year ago. It can sometimes be months or even years before the company is sure it has a winner—in other words, that people don’t just accept an offer but also use the card often and pay their bills. But “when we see an idea is working we put our foot very hard on the gas pedal,” Morris says. In addition to figuring out what offers are most profitable, Capital One has learned to avoid telemarketing. Capital One calls customers only when running small tests of certain new products or conducting a brand-recognition survey. The company actually markets some cards with the promise that customers will not receive telemarketing calls. Capital One eschews telemarketing because, for one thing, consumers loathe it. For another, it has learned that customers are far more likely to buy something when they call Capital One, rather than when Capital One calls them. So the company relies heavily on direct-mail campaigns and preps itself to sell intelligently to the some 1.5 million people who phone its call centers each week. The company has plenty of opportunities to sell by phone even without placing a call. To take maximum advantage of the opportunities, the company introduced a call-routing process in 1998 that taps into its huge databases on millions of customers nationwide. When a customer calls in, software tries to identify the caller even before answering the phone. The software also tries to figure out the reason for the call, based on the person’s history or other criteria, such as whether the person is late in making a payment. The software then sends the call, together with the person’s account information, to a Capital One representative who has expertise in the area the customer is likely calling about. The rep gets the information, plus suggested “cross-sell” opportunities, even before the phone begins to ring. When a customer calls to report a lost card, for instance, that’s a good chance to sell him some credit-card protection. At one of the company’s corporate campuses in Richmond, Va., Capital One call-center associates celebrate successful “cross-sells” every few minutes by reaching out to give a shake to one of the New Year’s Eve-style noisemakers placed between cubicles. The noisemakers make just enough of a rustling sound to signal to nearby colleagues that someone has scored. A subdued round of applause follows. Reactions to these sales were once rowdier, but they were toned down after too many customers on the other end of the phone began asking whether there was a party going on in the background. Capital One does exercise some restraint. On a recent day, a customer called to admit sheepishly that she had melted her credit card in her microwave oven. She had frozen the plastic card in a block of ice to keep herself from overspending, then put the whole thing in the microwave when she needed the card to do some holiday shopping. This was not the first such call to Capital One. Associates are trained to point those customers to the financial education section of the company’s Web site. Capital One also has softened its rules by empowering its collection agents to decide how best to handle certain customer situations. “There are some parameters [that these advisers must stick to], but you can’t make a rule for everything,” says Marjorie Connelly, executive vice president of credit-card operations at Capital One. The company reasons that if it can be helpful when a customer gets into a tough spot, and tailor its response to an individual’s situation, it’s far more likely to get its payment. Mostly, though, the company is plunging ahead as aggressively as it can. Recently, the company has applied its information-based strategy to the tricky art of collecting money. In the past, credit-card companies followed the same scenario: When a customer was 15 days behind on a payment, he would get a letter. After 30 days, he would get a follow-up phone call. But would a customer be more likely to pay if he got a letter five days after missing a payment? Or what if he got a phone call before the letter? Capital One started playing with lots of different combinations—even experimenting with whether customers reacted better when they received a friendly call or when the caller had an all-business attitude. Among other things, Capital One learned that some customers are less likely to pay if called too much. Capital One even tests employees’ reactions to the food in its cafeterias before it rolls out new menus nationwide. The company has made extensive use of its facility in Tampa, Fla., to gauge employee reactions to such unusual corporate amenities as a nail salon. The company applies its “test and learn” methods to its philanthropic activities, too. Executives hypothesized that employees would support traditional corporate contributions, such as to local ballets, but a survey found that most employees wanted charitable dollars spent on “kids at risk.” So Capital One chose, among other things, to support after-school programs that help schoolchildren with their homework and feed them a hot meal before sending them home. The company is careful to follow up on its contributions by attempting to assess whether the programs have helped to boost grades or lower dropout rates. So far, results have been positive. But if there is one thing that’s sure about Capital One, it’s that the company will keep testing.
|
