Today's chief challenge for banks is top-line growth: Between 1993 and 2000 the top 30 U.S. retail banks reported an average 14% growth in deposits, but less than 1% of it was organic—the rest came from acquisitions.

Why is growth so hard? First, because of market saturation. These days more than 90% of families have a checking account. Just about everyone who can use a mortgage has one.

The second problem lies in some banks' addiction to the last decade's challenge: managing their efficiency ratio, the measure of expenses as a proportion of fee income. They need to face this decade's imperative: investing in customers. Those that fail to do so will also fail to grow.

Our firm has taken a look at costs as a percentage of retail deposits at the largest banks. This investment ratio provides a handy gauge of how much banks spend managing call centers and branches, running marketing programs, and so on; in other words, it is a proxy for how much banks invest in each customer. We have found that retail banks with a ratio above 5% achieve top-line growth, while those below the threshold do not.

Between 1997 and 2002 the top 50 U.S. bank companies cut their efficiency ratios by an average of 1.47% just to stay in the retail banking game. Some still view cost control as their top priority and are paring costs so low they have no money to spend on customers.

Why does this matter? First of all, customers are getting choosier. The attention to service shown by nonbanks such as FedEx Corp. and Dell Computer Corp. has raised customers' expectations. These days, banks simply cannot get away with long lines at counters that close at 3 p.m.

Banks that do not invest are doing little to build customer relationships. Sure, their operating margins are as healthy as those of their high-spending competitors. PNC Financial Services Group Inc., a low investor, has an efficiency ratio of 57.7%, while at Wells Fargo & Co., a high investor, the figure is 56.3%. But PNC and other banks that underinvest in customers fail to achieve organic growth. Between 1999 and 2001 below-the-threshold banks achieved only 0.4% average organic growth.

Unfortunately, the cheapskates have not noticed that they are cutting their way to disaster. Any bank whose costs fall below 5% of the value of deposits will struggle to hang on to its customers over the coming years. Acceptable customer service simply costs more than that.

Not only do banks that underinvest in customers grow more slowly, but they also deliver lower returns. Our study shows that banks with investment ratios below 5% typically earn only 73% of the returns of those above the threshold.

On the other hand, banks whose investment ratios exceed 5% find they can build their top line. Between 1999 and 2001 companies such as Wells Fargo, which has a ratio of 6.7%, and BB&T Corp., where the ratio is 5.3%, had an average organic deposit growth of 7.1%.

Banks that invest above the threshold have found different ways to appeal to customers. Some, like the retail units at Wells and Wachovia Corp., excel at deepening their customer relationships. They offer a broader range of products and back them up with advisory services.

Wells sells more than four products to each household it serves and has set an ambitious 2006 target of eight products per household. (Low-investing competitors appear to sell an average of around 2.5.) This depth in customer relationships shows up in the top line: Wells' annual revenues equal 9.3% of deposits, compared with 5.5% for an average below-the-threshold bank.

Another set of above-the-threshold companies, including Commerce Bancorp Inc. of Cherry Hill, N.J., and North Fork Bancorp Inc., are luring customers away from competitors by offering premium convenience and service.

Between 1998 and 2001 Commerce nearly doubled its branches and extended its hours; some branches serve customers a whopping 78 hours a week. It puts greeters at entrances so that customers can quickly determine which service desk they need, and it invested $10 million in "penny arcades" that convert the contents of piggy banks into crisp notes. It even provides dog biscuits for its customers' pooches.

Competitors may scoff, but these strategies are steadily eroding other banks' market share. Commerce increased its deposits by 38% in 2001, compared with industry growth of 8%.

Another winner, North Fork, has invested heavily in Manhattan, where it opened 20 branches between 2000 and 2002 and achieved 170% growth in fee income.

What these companies recognize—and what below-the-threshold ones appear to miss—is this: There is no room for a rock-bottom basic offering in branch banking. Any player of a reasonable size can offer bread-and-butter products like checking accounts and mortgages at very low prices. But bread and butter will not attract customers' dollars. That takes cream, in the form of insurance and investment products.

Low-spending banks need to start investing in customer relationships. They need to put investment ratios on their radar screens, alongside efficiency ratios. Where they should invest depends on what their customers value most, whether it is improved service, extra branches, more products, expert advice, or new channels.

If these banks fail to spend more on their clients, their customers may not stick around. Instead, they may cross the parking lot to the banks that provide advisers, sofas, and treats for their dogs.

Mr. Aboaf is a partner and Mr. Bindra a manager at the New York office of Bain & Co.