Supermarket banks started the decade as a novelty; six years and more than 3,000 branches later they are one of the most visible signs of the changes sweeping the personal financial services industry. In California, Wells Fargo has built a supermarket network of more than 700 branches and in-store kiosks; Bank of America has 200 supermarket branches. Smaller networks have been built by, among others, Fifth Third, NCBC, and NationsBank, while SunTrust, First Chicago, Star Bank, and Michigan National Bank, along with a host of others, are experimenting with the concept. By 2000, one in ten US bank branches will be in supermarkets.

The reasons for the boom are obvious. Supermarket banks offer clear benefits to supermarkets and customers, as well as to banks. The supermarket gets rent from the bank and increased customer loyalty, while customers can bank at a branch that is open seven days a week until 9pm, and do their shopping at the same time. But the trend is mainly driven by what the banks get out of it. Compared with a traditional branch network, a well thought-out supermarket network has incredible economics (Exhibit 1). A high-performing supermarket bank employee can sell 50 percent more checking accounts, 40 percent more consumer loans, and 100 percent more credit cards per month than the average branch banker (although sales of high-margin investment products are 30 percent lower in supermarket branches). Balance levels, margins, and fees are generally comparable.

A supermarket branch also requires less commitment in up-front cash and also enjoys lower operating costs. A traditional branch can cost up to $1.5 million to construct, while a supermarket branch typically costs $200,000. The total expense of an in-store branch is about two-thirds that of a traditional branch.

Yet these seductive economics are far from given. Despite the potential, many banks lose money on supermarket networks because they do not understand how supermarket banking fits into their overall retail network strategy. Many banks have located branches in supermarkets that generate too little traffic to support a large number of new banking customers; to break even there need to be a minimum of 20,000 to 25,000 shoppers visiting at least once a week. Banks that have opened branches in stores with traffic below this threshold have often failed.

Other banks have overpaid for supermarket space, because initially they calculated a "reasonable" rent per square foot based on a percentage of their rent for a traditional branch. But supermarket economics are fundamentally different from bank economics. One early entrant into supermarket banking paid so much rent that the branch was the supermarket’s highest-margin item.

Perhaps the biggest mistake a bank can make is not recognizing that supermarket banking is different from traditional branch banking and needs to be managed differently. Banks often maintain a traditional, reactive sales culture in supermarket branches, and fail to create the aggressive sales strategy that would translate their exposure to a large number of new customers into actual sales that would support the cost of the additional branch. Without it, supermarket branches tend to open mainly low-balance, secondary, convenience checking accounts and rarely secure the customer’s total banking relationship.

Avoiding these mistakes and realizing the potential of supermarket banking is a complex and difficult task that requires thorough knowledge of the market, an objective examination of a bank’s overall retail strategy, and an understanding of the economic and management challenges involved in an in-store branch network.

What is my strategy?

Banks have experimented with four types of in-store format over the past decade: a traditional branch transplanted into a supermarket, a scaled-down version of a traditional branch, a leaner and more aggressive format, and the "one-man band."

The traditional format has been less successful because of its passive sales approach and high costs. The transplanted conventional branch, used by almost all early entrants into supermarket banking, was, in almost every case, a dismal failure. No bank uses this model today.

The scaled-down in-store branch is a reduced version of the traditional branch, offering limited services and employing a smaller staff. This model, still used by several banks including NationsBank in Atlanta, has had mixed success. Its attractive cost structure has enabled those first into a market to open more outlets, and therefore acquire new customers. But most banks using this format have copied the passive marketing approach of a traditional branch and have found it difficult to generate additional sales or to build lasting relationships. This effect reportedly contributed to BankSouth’s decision to be acquired by NationsBank. BankSouth had a strong presence in its market, but no effective channel for increasing its share of customers’ business.

The leaner, aggressive retail format and the one-man band have been used successfully by many banks, particularly those pursuing supermarket banking as a main strategic thrust. The aggressive retailer offers full-service banking through a highly proactive salesforce, an approach which has proved effective in acquiring customers and generating cross-sales. This way of operating reduces transaction costs by directing routine inquiries to automatic teller machines (ATMs). It also serves as a sales training ground for the next generation of bankers.

The one-man band is similar to the aggressive retailer, but costs even less. It consists of a kiosk staffed by one or two bankers and equipped with an ATM and a fold-down desk. Wells Fargo’s one-man "banking center" consists of a unit of 27 to 36 square feet with a laptop computer for opening accounts, a card reader for changing personal identification numbers, a combination photocopier/fax machine for paperwork, a telephone, and an ATM. Instead of a desk, the banker writes on a foot-long shelf. He or she roams the store seeking customers and diverts almost all transactions and servicing to the ATM. Some banks have built supermarket networks combining aggressive-retailer and one-man band formats. For the cost of five traditional branches, Wells Fargo claims it can operate four aggressive-retailer branches and eight supermarket kiosks and still save $1 million a year.

Whatever the format, supermarket banking is generally most successful for banks that want to focus on younger customers, especially professionals and dual-career couples (who are attracted to the channel’s convenience), or on the mass market, which can be served cost-effectively through supermarket branches—and when it is selling simple, competitively priced services (Exhibit 2). This customer and product focus can be either the bank’s traditional approach, as it was at BankSouth, or the result of an explicit strategic shift, as at NCBC (see the boxed insert).

Supermarket banking fits less well with banks that target older, wealthier customers and promote a full-service, personal relationship emphasizing complex products. Neither Wachovia’s relationship strategy, which uses personal bankers, nor Northern Trust’s focus on affluent customers and trust services lend themselves to the aggressive sales approach of the supermarket outlet.

But where supermarket banking is a good fit, success or failure will turn on the specifics of each market. What are the supermarkets like? How big are they? How many customers do they have and how often do they visit? What are their rents? The ideal store has 60,000 to 75,000 square feet and is new enough to make costly refurbishment unnecessary. Traffic should be sufficient to provide ample new customers, and shoppers should visit the store often enough to conduct their banking there—the 20,000 to 25,000 weekly shoppers mentioned earlier. Most banks have negotiated a long-term lease with a set rental rate; a monthly rent of $70 to $90 per square foot is average.

The availability of good supermarket partners is critical. Wells Fargo and Bank of America so dominate leading Californian markets that new entrants to in-store banking are almost excluded. Many attractive partners in other markets are already taken; in Atlanta, Kroger is paired with NationsBank and Publix with SunTrust. Banks must also decide whether an exclusive relationship is desirable. Bank of America, for example, has developed a single partnership with American Stores, which operates Lucky stores in California and Jewel/Osco in Illinois. A Bank of America branch or ATM is to be found in most Lucky stores and in every Jewel/Osco store. Wells Fargo, on the other hand, operates branches in four grocery chains in California alone—Ralph’s, Vons, Safeway, and Albertsons. It believes multiple relationships enable it to select the best locations and open more outlets.

An important point in determining the success of in-store banking is the bank’s market power compared with that of potential partners, because it determines how the value created will be shared (Exhibit 3). Dominant banks stand to gain less from increased customer traffic and will be interested mainly in reducing branch costs. If they team up with a dominant supermarket, both will share the profits in this "negotiation among giants." If the supermarket environment is more fragmented, the dominant bank can choose from a number of potential partners all eager to share its traffic and customer loyalty, and which therefore charge minimal rent. The downside of this "banker’s choice" market is that the bank has to form partnerships with several supermarkets to cover the market fully.

Less dominant banks will gain more from the supermarket’s customer traffic. Teaming up with a dominant supermarket creates "grocer’s choice," which means that the supermarket is likely to negotiate most of the value for itself, charging high rent or perhaps a percentage of the branch’s profits. This situation is common in Europe, and has contributed to the slower adoption of in-store banking there. When both bank and supermarket each have a minor market share, a "marriage of minnows" is created which might give neither significant advantage.

The different market situations for supermarkets and banks in Phoenix and San Antonio, for example, have greatly affected the growth of supermarket banking in the two cities (Exhibit 4). In Phoenix, Fry’s is the leading grocer, but several other supermarkets have a sizable share of the market. Bank of America, BancOne, and Wells Fargo likewise all have relatively large local market shares. Given this "negotiation among giants," in-store banking has been successful in Phoenix. In San Antonio, on the other hand, one supermarket, H-E-B, dominates, but no one bank does. Hence no bank can negotiate a favorable partnership with H-E-B—a case of "grocer’s choice." Penetration of supermarket banking in San Antonio is therefore low. H-E-B could feasibly open its own supermarket branches—without a bank partner—to capture the full margin on bank products.

What is the role?

Figuring out each supermarket branch’s role in the retail network is critical, because the convenience of the overall network matters more than the location of any one branch (many customers open an account at a given bank in order to conduct transactions in more than one branch—whether supermarket or traditional—within their area). In markets in which a bank performs well and has good growth prospects, low-cost, in-store branches can supplement existing branches by generating additional growth. Bank of America has used this strategy to open a network of in-store branches, without closing traditional outlets—although it is not clear if the bank’s markets offer sufficient growth potential to support its increased cost structure, or whether it will have to close some traditional branches to achieve its profit goals.

In areas that boast neither good performance nor growth opportunities, in-store branches can be less costly replacements for conventional ones. They will tend to focus on existing customers rather than on acquiring new ones. Before its acquisition of Wells Fargo, First Interstate Bank planned to close 300 of its 600 traditional branches in low-to-medium performing markets and replace them with small, highly productive supermarket branches.

Supermarket branches can also be a route into new markets (although it is tough to build a convenient network without at least a few traditional branches, particularly a "downtown hub" branch to serve business customers). Bank of America provides the most striking example of this approach. It created its Midwest Retail Division to enter the Chicago area by putting banks into all Jewel/Osco stores, but opened no traditional branches (its multichannel network does, however, include ATMs, a phone center, and PC banking). National Commerce Bancorp also used this tactic successfully to expand from its home market in Tennessee into North Carolina and Virginia.

Because it is unique, supermarket banking needs to occupy a distinct place on a bank’s organizational chart and in its management structure: one bank groups the manager of its in-store network with the manager of its traditional branch network under the head of distribution. It should also have a separate sales and marketing organization. In-store bankers need to walk supermarket aisles and sell their services to customers, not just conduct transactions back in their kiosks. They should emphasize the full array of higher-margin, complex products, and take advantage of the extended opening hours to sell to family decision makers during their non-working hours. Constant, varied promotions are important.

To support the aggressive sales culture, supermarket branches need a separate recruitment and training program. Staff are typically young—the average age of one bank’s in-store branch managers is 25—and retailing oriented. Sales skills are all-important, banking skills secondary (a strong referral and support structure should be in place to handle complex transactions). Measurement systems too should be tailored to the environment. Many banks pay a commission for new accounts (usually $5 each). Sales targets are usually higher than they are in traditional branches, and results may be posted on a "leader board" to keep bankers motivated.

While banks will have to balance investment in physical outlets with the development of remote distribution channels such as telephone- or PC-banking—the superior economics of which will make them increasingly important as customer acceptance grows—further retail combinations are likely. Any outlet that has a high throughput of customers (who also make frequent return visits), and in which customers spend enough time for a banker to make a sales pitch offers potential. Look out for banks in discount stores such as Target or Kmart, in restaurants such as Burger King or Pizza Hut, and even in Blockbuster Video stores.