Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Retail Banking1 FRANCES X. FREI Harvard University PATRICK T. MARKER LARRY W. HUNTER University of Pennsylvania INTRODUCTION How does a retail bank innovate? According to the view in traditional inno- vation literature, organizations innovate by getting new and/or improved products to market. However, in a service industry like retail banking, the product is the process of serving customers. Thus, innovation in retail banking lies more in process and organizational changes than in new product development in a tradi- tional sense. This chapter reviews a multiyear research effort on innovation and efficiency in retail banking and discusses the means by which innovation occurs in retail banks. It also examines factors that make one institution better at inno- vating than another. The chapter draws implications to the broader service sector. We conclude that there is simply no "silver bullet" no single set of man- agement practices, capital investments, and strategies that lead to success. Rather, it appears that the "devil" is truly in the details. The key to efficiency in this industry appears to be the alignment of technology, human resources manage- ment, and capital investments with an appropriate production "technology." To achieve this alignment, banks need to invest in a cadre of "organizational archi- tects" that are capable of integrating these varied pieces together to form a coher- ent structure. The biggest challenge facing retail banks with respect to efficient and effec- tive innovation lies in the management of the "New Age Industrial Engineers" iThis research was supported by the Wharton Financial Institutions Center through a grant from the Sloan Foundation and the National Science Foundation's Transformation to Quality Organizations, Grant SBR-9514886. 179
80 U.S. INDUSTRYIN2000 that must combine technological knowledge with process design to create the delivery systems of the future. THE INNOVATION CHALLENGE IN FINANCIAL SERVICES Financial services are a huge and critical sector of the U.S. economy, com- prising over 4 percent of the U.S. Gross Domestic Product (GDP) and employing over 5.4 million people more than double the combined number of people em- ployed in the manufacture of apparel, automobiles, computers, pharmaceuticals, and steel.2 While impressive, these numbers belie the much larger role that this industry plays in the economy (Herring and Santomero, 1991~. Financial ser- vices firms provide the payment services and financial products that enable house- holds and firms to participate in the broader economy. By offering vehicles for investment of savings, extension of credit, and risk management, they fuel mod- ern capitalistic society. While the essential functions performed by the industry the provision of payment services, and facilitation of the allocation of economic resources over time and space have remained relatively constant over the past several decades, the structure of the industry has altered dramatically. Liberalized domestic regu- lation, intensified international competition, rapid innovations in new financial instruments, and the explosive growth in information technology are fueling this change. Against this backdrop, managers and workers face intensified pressure to improve productivity and financial performance. Competition has created a fast-paced industry where firms must adapt and innovate to survive. Given the increasing competition in the financial services industry and rapid technological evolution, how do retail consumer banks innovate to meet these challenges? This chapter attempts to answer this question by considering general trends in retail banking and by describing a detailed field study at a major U.S. retail consumer bank. We discuss the forces that are driving retail banks' need to innovate and describe the means by which banks innovate. In the process we discuss what constitutes efficient and effective innovation in banking. After all, not all innovation is good, and even if the innovation is a good idea, the costs of execution can substantially exceed the benefits. The Changing World of Retail Banking Nowhere are the changes sweeping the financial services industry more strongly felt than in retail consumer financial services. Once the sole domain of 2Comparison based on average 1991 data reported by the U.S. Bureau of Labor Statistics, Employ- ment and Earnings Report, March 1992. Data for the financial services industry includes SIC codes 60-64 and 67. Data for the apparel, automobile, computer, pharmaceutical, and steel industries in- clude SIC codes 239 (less 23), 371, 357, 283, 331, and 332.
RETAIL BANKING TABLE 1 Changes in the U.S. Banking Industry 1979-1994 181 Item 1979 1994 Total number of banking organizations 12,463 7,926 No. of smallbar~ks 10,014 5,636 Real industry gross total assets (tnllions of 1994 dollars) 3.26 4.02 Industry assets in megabanks (percent of total) 9.4% 18.8% Industry assets in small banks (percent of total) 13.9% 7.0% Total loans and leases (trillions of 1994 dollars) 1.50 2.36 Loans made to consumers (percent of total) 19.9% 20.6% Total number of employees 1,396,970 1,489,171 Number of automated teller machines 13,800 109,080 Real cost (1994 dollars) of processing a paper check 0.0199 0.0253 Real cost (1994 dollars) of art electronic deposit 0.0910 0.0138 Note: A "megabank" in this table is a bank with over $100 billion in assets in real 1994 dollars. A "small" bank is one with assets under $100 million in 1994 real dollars. Source: Berger et al. (1995). the retail bank, various non-bank competitors including mutual funds and bro- kerage firms are increasingly offering competing services, eroding the market share of the traditional retail banking sector. Consider the changes depicted in Table 1. It is also investing heavily in new information technology primarily to facilitate new electronic means of transactions, which hold a cost advantage over traditional paper-based banking. The major forces for these changes will be described in detail in the next section, but a quick glance at Figure 1 confirms that increased competition from other players in the financial services industry continues to erode the market- share of banks. This competition, along with the explosive changes in informa- tion technology (IT) and changes in consumer demand, fuels the need for banks to innovate in products, services, and delivery channels. THE FORCES OF CHANGE IN RETAIL BANKING Various forces are driving change in retail banking, but the principal ones are regulatory changes, technological innovation, and changing consumer demand. Regulatory Change and Consolidation As shown in Table 1, the retail banking industry is undergoing a period of rapid consolidation as well as expansion into non-traditional banking products and services. Between 1979 and 1994, approximately 5000 banking organiza- tions were taken over by other depository institutions. Why? First, regulations restricting interstate banking and the broadening of product lines of the banks continue to weaken. Changes regarding reserve limits, bank
182 100 80 60 o cow it' 40 o 1980 1985 1990 1995 Other Mutual funds Insurance and Pension funds Stocks Bonds Bank deposits FIGURE 1 Share of U.S. consumer financial assets, 1980-1995. Source: Federal Reserve data, reproduced in Council on Financial Competition (1996). U.S. INDUSTRYIN2000 powers, geographic restrictions, and the Glass-Steagall Act restrictions on prod- uct offerings have all fueled merger activity.3 Banks are also responding quickly to the removal of limits on interstate banking activities, as shown in Table 2. Similarly, the relaxation of the Glass-Steagall restrictions on bank holding companies has permitted banks to merge across product lines. Bank holding TABLE 2 Changes in the Geographic Focus of the U.S. Banking Industry 1979-1994 Item 1979 1989 1994 Total national banking assets (%) legally accessible from a typical U.S. state Typical state's banking assets controlled by out-of-state multibank holding companies Source: Berger et al. (1995). 6.5% 29.0% 69.4% 2.1% 18.9% 27.9% 3See Berger et al. (1995) for a detailed discussion of these regulatory changes.
RETAIL BANKING 183 companies are increasingly purchasing mutual fund companies, brokerage houses, and insurance firms in order to offer a full spectrum of financial products to their customers. These cross-industry acquisitions are aimed at stemming the contin- ued erosion of market share depicted in Figure 1. The driving force in every bank is "share of wallet" the desire to attract and retain more and more of a con sumer's financial business. Do these mergers work? At present, the evidence is quite mixed in terms of both cost reduction and profit efficiency.4 In terms of shareholder value, recent research suggests that these mergers have tended to destroy, not enhance, value as shown in Figure 2. One major explanation for retail banking's consolidation is the desire to have sufficient size to exploit scale economies in transaction processing and scope economies in cross-selling multiple financial products to a household. However, numerous studies of efficiency in the banking industry show that neither scale nor scope efficiency is the main cause of inefficiency. Summarizing this research, Berger et al. (1993) focus on a measure X-efficiency that isolates all technical and allocative efficiencies of individual firms that are not dependent on scope or scale. That is, X-efficiency captures how well management is aligning technol- ogy, human resources, and other assets to produce a given level of output. They note, "The one result upon which there is virtual consensus is that X-efficiency differences across banks are relatively large and dominate scale and scope effi- ciencies." Other results, such as those reported by Fned et al. (1993) in the context of credit unions, add additional weight to the importance of X-efficiency by provid- ing evidence that it is a dominant factor in both large and small institutions. Based on this evidence, it is clear that scale and scope economies are not the driving factor in explaining firm-level efficiency and the driving force behind mergers. Value destroying Value creating Borderline FIGURE 2 Shareholder value analysis of bank mergers and acquisitions 1983-1988. Source: D.C. Cates (1991). 4Some studies, such as Shaffer (1993) and Akhavein et al. (1997), show that banks can obtain lower costs and increased profits, while others (Rhoades, 1993; Peristiani, 1997) show little to no post- merger gains.
84 U.S. INDUSTRYIN2000 Summarizing the problems of inefficiency in this industry, Berger et al. (1993) state: Our results suggest that inefficiencies in U.S. banking are quite large the in- dustry appears to lose about half of its potential variable profits to inefficiency. Not surprisingly, technical inefficiencies dominate allocative inefficiencies, sug- gesting that banks are not particularly poor at choosing input and output plans, but rather are poor at carrying out these plans. What then drives the consolidation of the industry? When questioned on their strategic response to increased competition, bank directors stated that acqui- sitions were the most important method for overcoming competitive threats and positioning themselves for the future (see Figure 3~. Thus, much of the consoli- dation can be viewed as a strategic response to an acceleration of change in the industry. Many bankers are worried about firms like Microsoft entering the bank- ing business. To face this competition, they feel that they must extend both scale and scope in order to compete in the future. Obviously, not all banks that merge or acquire other institutions are achiev- ing negative results. Just like the inefficiencies described above, there is a distn- bution of talent when it comes to consolidation. In a recent paper, Singh and Zollo (1997) discuss the role of organizational experience and learning in the bank acquisition process. Summarizing their results, the authors state: "The prob- ability of a high level of integration [of banks] is strongly determined by the degree to which the acquirer has codified its understanding of how to accomplish this extremely complex and relatively infrequent task." Thus, the acquisition process itself can be viewed as a major source of innovation in banking. `~ 100 . _ .~ c' :~ c' it c' ~o 80 60 40 20 Acquirebank Focus on Focus on Merge with Exit business or thrift product market a bank lines segment Strategy FIGURE 3 Bank director's response to the following question: What will you most likely do to overcome competitive threats and better position yourself for the future? Note: Over 200 bankers were surveyed. Source: Towers Perrin's 1994 survey (1996).
RETAIL BANKING 185 Mergers and acquisitions, therefore, are a powerful force of change in the banking industry, impacting not only the geographic scope and product variety of the organization, but also affecting the underlying technological and manage- rial infrastructures of the banks. For the foreseeable future, consolidation will continue to position the organizations against current and future players in the marketplace. Technological Innovation Technology plays a key role in the performance of banks. Large U.S. banks spend approximately 20 percent of non-interest expense on information technol- ogy, and this investment shows no signs of abating. Even with these large invest- ments, it is still difficult to ascertain the payoffs associated with these projects. In manufacturing, recent studies have found large payoffs in IT investments, in terms of both equipment and personnel (Brynjolfsson and Hilt, 1993; Lichtenberg, 1995~. For example, Lichtenberg (1995) states that "...the estimated marginal rate of substitution between IT and non-IT employees, evaluated at the sample mean, is six: one IS employee can substitute for six non-IT employees without affecting output." Unfortunately, similar results for financial services are not available. For example, in the recent study by the National Research Council (1994) on IT in services, the problem in the context of banking is summarized as follows: Neither approach [for productivity measurement] is able to account for improve- ments in the quality of service offered to customers or for the availability of a much wider array of banking services. For example, the speed with which the processing of a loan application is completed is an indicator of service that is important to the applicant, as is the 24-hour availability through automated teller machines (ATMs) of many deposit and withdrawal services previously acces- sible only during bank hours. Neither of these services is captured as higher banking output at the macroeconomic level. While hard-and-fast data are not yet available, many believe that financial services are at the brink of major performance improvements due to technology. However, this will not occur in the traditional "back-office" functions such as check processing. Rather, the performance improvements will result from the integration of front- and back-office functions that is, in integrating business processes. Roach (1993) points out that the consolidation of back-office opera- tions is due in large part to scale economies resulting from to IT investments but that these investments are becoming increasingly difficult to find. However, he states that "...new productivity opportunities are now spreading rapidly across the sales function of the service sector...." It is precisely in these front-office func- tions that major investments will occur. Philip Kotler (as cited in Pine, 1993) states this trend clearly:
86 U.S. INDUSTRYIN2000 Instead of viewing the bank as an assembly line provider of standardized ser vices, the bank can be viewed as a job shop with flexible production capabilities. At the heart of the bank would be a comprehensive customer database and a product profit database. The bank would be able to identify all the services used by any customer, the profit (or loss) on these services and the potentially profit able services which may be proposed to that customer.... This movement away from mass marketing, mass production, and mass distribution is widespread throughout the financial services industry. Technological innovation in the retail banking industry has been spurred on by the forces described by Kotler, particularly in terms of new distribution chan- nel systems, such as PC banking. As the industry has provided more ways for consumers to access their accounts, they have added significant costs to each institution. A need to combat these costs resulted in a major effort to reduce cots in back-office operations through automation contributing to productivity im- provements in functions and the processing of loan applications. Now, after add- ing significant costs through added distribution channels and cutting as much as possible in the back-office, banks have realized that the key to profitability is through revenue enhancement. Banks are now forced to consider new ways to drive revenue through their distribution system. The most common way to do so is to try to increase the share of the customer's wallet. As explained above, the share of wallet is the portion of a customer's entire financial relationship that any particular bank has with the customer. The prevailing hypothesis is that the more products that a customer has with the bank, the cheaper it is to serve them per product and the more diffi- cult it would be for the customer to switch to another bank. The primary revenue-enhancing innovations occurring today are in platform automation, i.e., the automation of the functions performed by front-line employ- ees, for branch and phone center employees, and in the newest distribution chan- nel, PC banking. While these innovations have aspects in common, they serve different needs in the distribution strategy of retail banks. Platform automation is the retail banking industry's first major attempt at giving employees a single view of the customer. Prior to this innovation, it was not possible for an employee to view the entire customer relationship at one time. Why is this important? First, a single view lets the employees understand how important a customer is based on their portfolio of products rather than on their current checking account balance. If hidden behind that low checking balance is a series of CDs and a home equity loan, for example, then the employee may want to think twice before refusing to waive a small fee associated with the check- ing account. However, although the concept of bringing together all of a customer's relationships with the bank is quite simple, in reality it has proven to be an extremely difficult task. Retail banks collect and process information by product and transaction, not by customer. While it is quite easy to access all the information on checking
RETAIL BANKING 187 account customers or on credit card customers, for example, taking a slice of the data per customer is technologically difficult. Virtually every bank has been faced with this same problem. Legacy systems (i.e., 1970s-style data processing software) were built with transaction processing per product in mind. Banks have excellent check processing systems, but such systems do not easily interact with the mortgage processing system, for example. Moreover, the data are scattered among a variety of systems and locales. It is quite common to have credit card processing in a different state from the rest of the retail bank, so that bringing this data together is a massive undertaking. PC banking represents a new distribution channel and an area for significant technological innovation. With this new channel, there are many alternatives available to each bank, and with these alternatives come managerial decisions regarding alliances, outsourcing, new product development, and a host of other critical factors that will influence future profitability. At the surface, one could consider the PC channel similar to the phone center in that a customer is simply contacting the bank remotely, in one case over the phone and in the other by the PC. The major difference between the channels comes in the variety of ways that a bank can offer PC banking and in the implications resulting in each model. We describe the four most common PC banking models below in order to demon- strate the variety of alliances and outsourcing practices as well as to discuss the implications of each in terms of potential loyalty and increased share of wallet. Coincident with the retail banking industry moving from cost-savings inno- vation to revenue-enhancing innovation is the move from in-house development to outsourcing and alliances. While there are many arguments favoring this shift, including the most common view that banks are not software companies and should not be developing these systems in house, it remains to be seen if this shift will loosen the bank's stronghold as the predominant financial intermediary. As payment systems in the United States catch up to the rest of the world in terms of the ability to have end-to-end electronic processing, it is not clear where the prof- its will be made. By making choices today in terms of platform automation and PC banking models, banks are making explicit choices about where they see them- selves in the future. The Changing Consumer The final, and perhaps the most important, force of change in the banking industry is the rapid evolution of consumer wants and desires. Consumers are demanding anytime-anywhere delivery of financial services, along with an in- creased variety in deposit and investment products. Consider first the desire for greater product diversity. Whereas Fidelity In- vestment and Merrill Lynch both offer over 100 different choices for mutual funds, the typical bank offers 17.5 As a result, banks continue to lose market 5"Mutual Fund Review," Wall Street Journal, April 1996.
188 U.S. INDUSTRYIN2000 share (see Figure l). Choice of demand deposit accounts with a desired fee struc- ture, along with the advent of new investment vehicles such as index funds, all fuel the banking customer's desire for new and better financial products. In addition, consumers are moving away from the use of checks to other financial products, albeit slowly (see Figure 4~. Consumers are also demanding variety of delivery channels available for their use (Table 3~. In spite of the assertion that branch delivery is dead, most consumers still frequent the branch. In fact, there has been a rise in the number of branches, including supermarket- based locations (called "in-store branches") and kiosk-like branches found in many shopping malls. And, as can be seen in Figure 5, this trend to open new physical sites seems likely to continue. Furthermore, it is the "mixed channel consumer" one who frequents multiple delivery points that is the norm in the industry (Figure 6~. Consumers are demanding and receiving a larger variety of traditional and new banking products and delivery systems. The question, however, is how banks capture the value generated by this increase in variety. At present, one needs to look only at the controversy surrounding consumers' resistance to paying fees for various ATM transactions to understand that this increase in variety may be det- rimental to a bank's profitability. Over decades, banks have invested heavily in ATM machines because of their cost advantage on a per-transaction basis (Table 4~. The traditional teller transaction is almost an order of magnitude more expen 100 80 60 40 20 O 1990 1991 1992 1993 1994 Credit card payments Debit card payments Checks issues FIGURE 4 Use of various payment instruments (millions of transactions). Source: Kennickell and Kwast (1997), Table 1.
RETAIL BANKING TABLE 3 Percent of U.S. Households Using Vanous Delivery Channels 189 Delivery Channel % of households In person/ branch visit Mail Phone Electronic transfer ATM Debit card Direct deposit Pre-authorized debit/ payment PC banking 86.7 57.4 26.0 17.6 34.4 19.6 59.6 23.6 3.7 Source: Kennickell and Kwast (1997), Table 2. 100 80 60 40 20 o ,~,i/,,,~ ,,.,,,,,, ,,,,,,,, ,,,,,,,,. ,,,,,,,, ,,,,,,,,. i,,,,,,,. ,,,,,,,,. ,,,,,,,,. ,,,,,,,,. an_ Percent of Banks [g Reduce the number of branches Open new in- store branches Remodel existing branches Open new full- service branches Open new kiosk branches ~1 Open new min branches FIGURE 5 Branch activities planned over the period 1995-1998. Source: Ernst and Young (1996), annual survey of major U.S. banks. 40 30 20 10 o 1. Percent of Households One channel two channels Three channels >= Four channels FIGURE 6 Percentage of U.S. households using various numbers of delivery channels. Source: Kennickell and Kwast (1997), Table 2.
90 U.S. INDUSTRYIN2000 TABLE 4 Comparison of Cost per Transaction for Various Delivery Channels Distnbution channel Cost per transaction Teller Telephone (human operator) Telephone (automated voice response unit) ATM $1.40 $1.00 $0.15 $0.40 Source: Oliver Wyman arid Company. sive than ATM and automated phone systems. This has led banks to attempt to change consumer behavior through the addition of fees (the "stick") and a variety of rebates (the "carrot". Despite these efforts, the total cost of serving certain customer segments has not changed significantly because customers, in the mean- time, have altered their own transaction behavior. Some customers even use ATMs more frequently: a typical college student may use the ATMs once a day, for one $20 bill! Changing customers' behavior to use ATMs instead of tellers, for example, or to use ATM's less frequently would, in theory, yield the great- est benefit to banks in terms of cutting costs. However, in practice, this change in behavior will be difficult to achieve, as evidenced by the recent customer uproar over increases in ATM fees. Thus, banks must continue to innovate in order to meet the changing needs and desires of the consumer. In particular, banks seek to leverage the develop- ments in information technology to create new products and services. At the same time, banks must develop new fee structures to shift consumers away from high-cost delivery systems. This blend of innovation and behavior change lies at the heart of the modern banking organization. We now turn to the innovation mechanisms banks use to meet these challenges. HOW DO BANKS INNOVATE? How does a retail consumer bank innovate? To begin to answer this question, consider two important developments in banking the emergence of the PC/elec- tronic delivery of financial services (a product innovation) and creating new dis- tribution channel designs (an organizational innovation). Both have had signifi- cant impact. In the case of PC banking, this innovation promises to revolutionize the cost structure of retail banking. However, such changes will not occur unless banking organizations can adapt their structure to exploit the new technology. Both technical and organizational innovation are crucial to retail banks. A Product Innovation: PC Banking Pushed by growing consumer demand and the fear of losing market share, banks are investing heavily in PC banking technology (Frei and Kalakota, 1997~.
RETAIL BANKING 191 In collaboration with hardware, software, telecommunications, and other compa- nies, banks are introducing new ways for consumers to access their account bal- ances, transfer funds, pay bills, and buy goods and services all without using cash, mailing a check, or leaving home. The four major approaches to home banking are, in historical order: . Proprietary Bank Dial-up Services. A home banking service, in combi- nation with a PC and modem, lets the bank become an electronic gateway to customers' accounts. This enables customers to transfer funds between accounts or pay bills directly to creditors' accounts. . Off-the-Shelf Home Finance Software. This category is essential in help- ing banks cement relationships with existing customers and gain new customers. Examples include Intuit's Quicken, Microsoft's Money, and Bank of America's MECA software. This software market is also attracting interest from banks because it has steady revenue streams through upgrades, updates, and the sale of related products and services. On-Line Services-Based. This category allows banks to set up retail branches on subscriber-based on-line services such as Prodigy, CompuServe, and America Online. World Wide Web-Based. This category allows banks to bypass subscriber- based online services and reach the customer' s browser directly through the World Wide Web. There are two great advantages: the flexibility to adapt to new online transaction processing models facilitated by electronic commerce, such as on-line bill paying, and the ability to eliminate the constricting intermediary or on-line service. . . In contrast to packaged software that offers a limited set of services, the on-line and Web-based approaches offer further opportunities. As consumers increasingly purchase items in cyberspace with credit cards, debit cards, and newer financial instruments such as electronic cash or electronic checks, they will need software products to manage these electronic transactions and recon- cile them with other off-line transactions. In the future, an increasing number of paper-based, manual financial tasks may be performed electronically on ma- chines such as PCs, hand-held digital computing devices, interactive televisions, and interactive telephones. Banking software must have the capability to facili- tate these tasks. Home Banking Using Bank's Proprietary Software On-line banking was introduced in the early 1980s when at least four major banks (Citibank, Chase Manhattan, Chemical, and Manufacturers Hanover) of- fered home banking services. Chemical introduced its Pronto home-banking ser- vices for individuals and Pronto Business Banker for small businesses in 1983. Its individual customers paid $12 per month for the dial-up service, which al
92 U.S. INDUSTRYIN2000 lowed them to maintain electronic checkbook registers and personal budgets, see account balances and activity (including cleared checks), transfer funds among checking and savings accounts, and best of all make electronic payments to some 17,000 merchants. In addition to home banking, users could obtain stock quotations for an additional per-minute charge. Two years later, Chemical teamed up with AT&T in a joint venture called Covidea meant to push the product through the second half of the decade. Despite the muscle of the two home-banking partners, Pronto failed to attract enough customers to break even and was aban- doned in 1989. Other banks had similar problems. Citicorp had a difficult time selling its personal computer-based home-banking system, dubbed Direct Access. Chase Manhattan had a PC banking service called Spectrum. Spectrum offered two tiers of service one costing $10 a month for private customers and another cost- ing $50 a month for business users, plus dial-up charges in each case. According to their brochure, business users paid more because they received additional ser- vices such as the ability to make money transfers and higher levels of security. Banc One had two products, Channel 2000 and Applause. Channel 2000 was a trial personal computer-based home-banking system available to about 200 customers that was well received. Applause, a personal computer-based home- banking system modeled after Channel 2000, attracted fewer than 1000 subscrib- ers. The trial was abandoned before the end of the decade, as the service could not attract the critical mass of about 5000 users that would let the bank break even. In each of the above instances, the banks discovered that it would be very difficult to attract enough customers to make a home-banking system pay for itself (in other words, to achieve economies of scale). Figure 7 describes a tradi- tional proprietary system of banking. On-line banking has been plagued by poor implementations from the early 1980s. Home-banking services lost too much from concept to reality. Many Consumer's PC Proprietary Bank's Software Interface Modem FIGURE 7 Proprietary software method for PC banking. Source: Fret and Kalakota (1997). Modem Bank Bank's Mainframe Computer
RETAIL BANKING Intuit's Quicken Personal Finance Software _ _ ~ _ ~ BANK BANK BANK I BANK Banks which allow on-line account access FIGURE 8 Banking with dial-up software. Source: Fret and Kalakota (1997). 193 / Local Point ~ f of Presence MODEM ~ (POP) \ Concentric Network ~;~ I Money ~ National Payment Processor (Intuit Services Corp.) I Bill Payment _ Automated \ ClearingHouse / -,,,,,~. systems evolved gradually, which often meant that consumers who initially used the service and left dissatisfied could not be coaxed back into using it again. Recently Citibank has revamped its Direct Access product, allowing con- sumers to dial in to Citibank's system and check their account balances, transfer money between accounts, pay bills electronically, review their Citibank credit card account, and buy and sell stock through Citicorp Investment Services. A1- though the underlying systems run in batch mode, Citibank has put together a middle-ware piece of software that makes consumers think that they are operat- ing in a real-time environment. While this can work in a setting where Citibank is not interacting with third-party systems, there are potential difficulties with this batch/real-time mix if Citibank offers outside products and services such as insur- ance products. In addition, because consumers are interacting directly with Citibank's system, they have no way of performing household budgeting func- tions on their financial data. Clearly, Citibank will need to either provide this function itself or provide easy interface to the popular personal finance packages. However, it is important to point out that the new Direct Access represents the first major improvement in proprietary software home banking in 15 years, which is demonstrated by their explosive growth from 40,000 subscribers to 190,000 in 1996. Banking with the PC Using Dial-Up Software The main companies that are working to develop home-banking software are Intuit, the maker of Quicken; Microsoft, the maker of Microsoft Money; Bank of America and NationsBank, who acquired Meca's Managing Your Money soft- ware from H&R Block; and ADP, which acquired Peachtree Software. Banking with third-party software means that there is an intermediary between the bank and the consumer. In fact, as can be seen in Figure 8, it is easy to imagine how
94 U.S. INDUSTRYIN2000 the banks can become back-end commodity providers in this system, with the third party controlling the customer interface. Banking with On-Line Services Although personal finance software allows people to manage their money, it only represents half of the equation. No matter which software package is used to manage accounts, information is managed twice: once by the consumer and once by the bank. If the consumer uses personal finance software, then both the con- sumer and the bank are responsible for maintaining systems that do not commu- nicate. For example, a consumer enters data once into his or her system and transfers this information to paper in the form of a check, only to have the bank then transfer it from paper back into electronic form. In the instance where an electronic check is issued, the systems that receive the information rarely com- municate automatically with bookkeeping systems. Unfortunately, off-the-shelf personal finance software cannot bridge the com- munications gap or reduce the duplication of effort described above. However, a few home-banking systems that can help are beginning to take hold. In combina- tion with a PC and modem, these home-banking services let the bank become an electronic gateway, reducing the monthly paper chase of bills and checks. The general structure of the on-line services banking architecture is shown in Figure 9. How to Innovate with PC Banking Although there is no clear choice as to the appropriate home-banking model, it is quite clear that very explicit trade-offs must be made. In addition to consid- ering control of the interface, security, speed of access, and convenience, banks must consider the level of customer support required for each model. Basically, intuit's Quicken as the Generic Front-end Bank's Customized Proprietary Front-end MODEM . . . _ _ _. . . / MODEM FIGURE 9 On-line services banking architecture. Source: Fret and Kalakota (1997). C tiOanlc = f ~: , ~ ~ ~ ) ~Chase/Chemical /~ it_ America Online Prodigy CompuServe ~ Bill Payment Payment Pn~ce~or 1
RETAIL BANKING 195 the larger the numbers of intermediaries, the higher the level of support the customer will need. Those banks that understand the technology, human re- source, and process issues will have a better chance of coming out ahead in this innovation. Thus, the fundamental challenges to innovation in PC banking are not tech- nological per se but arise from the complex set of organizational choices to imple- ment such a service for the consumer. Suppliers can provide not only the soft- ware needed to support a PC banking operation, but the back-office fulfillment processes as well. The basic innovation for the bank lies in its integration of these software and fulfillment processes to create the electronic banking service. To illustrate the fact that it is often organizational change that fuels innovation in banking, we now turn to an example of a bank that is in the process of re-creation. An Organizational Innovation: Re-Creating a Bank National Bank,6 one of the larger U.S. commercial banks, with branches in many states, has a retail banking arm that is in many respects typical of the indus- try. Our research team has spent the past year studying the process of innovation at National, tracking the implementation of a major redesign of the retail delivery system. Confronted by an increasingly competitive environment, National was chal- lenged with improving the cost-efficiency of its far-flung retail delivery system, comprising hundreds of branches. At the same time, National sought to trans- form these numerous retail branches into sales-focused financial stores: ones that concentrated more directly on the sale of financial products and services. Our account of the continuing process of redesign at National illustrates a number of the observations made earlier in this chapter. National's retail banking organization was quite decentralized. No single organizational unit in the bank had responsibility for retail operations. Rather, each of the major geographic areas served by the bank had its own management team. The challenge of redesigning the bank was heightened by the diversity across geographic areas. Some of the state-based operating divisions, and many of the branches, had been acquired from other banks and quickly folded into National, retaining many of their former employees and some of their technology and business processes. To carry out the redesign, therefore, National had to build from scratch a group responsible for its implementation. National assembled a re-engineering team of over 50 employees, drawn from a diverse set of geo- graphic areas and functional backgrounds, and charged this team with spearhead- ing the overhaul of the branch delivery system. Initially, the redesign at National focused around very basic business process re-engineering in the branches. Over a period of decades, a huge number of 6National Bank is a pseudonym, used to protect confidential information.
96 U.S. INDUSTRYIN2000 administrative functions had accumulated in the branch systems, so that branch managers and service representatives spent a considerable amount of time on these activities rather than in contact with customers. Further, most of the time spent with customers was centered on simple, transaction-oriented and basic ser- vicing of accounts rather than on activities that were thought to be likely to lead to sales opportunities. Leaders at National, recognizing these problems, engaged a leading consulting firm as a partner in the re-engineering of the branch system, and the consulting firm spent several months working with the implementation team to identify opportunities to streamline branch activities. The outcome of this partnership became known as the "pilot" redesign, and it was agreed that the redesign should be tested in a few small market areas before being rolled out across the bank more broadly. From the start, both the consultants and the team conceived the redesign to require broad, systematic change. Effective innovation therefore required the participation of virtually all the functional areas within the bank, from informa- tion systems to marketing to human resources, with each of these areas repre- sented on the implementation team. Anchoring the redesign was the streamlining of branch processes and the relocation of many of the administrative tasks and routine servicing of accounts to central locations outside the branch. To take one simple example, incoming telephone calls from customers were re-routed so that phones in the branch did not ring; rather, customers calling National and dialing the same number they had always used to contact the branch, would now find their calls routed to a central call center. The innovation also required redesign of the physical layout of the branches. A goal of the redesign was to encourage more customers to use automatic teller machines and telephones for routine transactions. Customers entering the rede- signed branch, therefore, were to be greeted by an ATM, an available telephone, and a bank employee ready to instruct them in the use of these technologies. The customer would be directed toward a teller or a service representative only if the customer insisted or when such personal attention was clearly necessary, for ex- ample, to deposit cash, to access a safe deposit box, or to meet with a sales repre- sentative about purchasing a product or service. These technological innovations, along with the redirection of customers to alternative delivery channels, were intended to make operations more efficient. As an example of the expected efficiencies, early projections by the consulting firm, although later shown to be overly optimistic, envisioned a 65 percent de- crease in the number of tellers required in the branch system. Over time, it was hoped that many customers would cease to rely on the branch and its employ- ees for routine transactions and services. The re-engineering was also expected to transform service employees into sales personnel, allowing them to concen- trate their efforts on activities that had potentially higher added value, such as customized transactions and the provision of financial advice coupled with sales efforts.
RETAIL BANKING 197 A clear requirement for effective innovation at National, then, was the par- ticipation not simply of the employees but also of the customers in the new ser- vice processes. In its design, National elected not to pursue some of the more notorious routes favored by other banks (such as charging fees to see tellers), but to lead customers somewhat more gently, by making customer relations a key feature of the redesigned retail bank. The redesign created a customer relations manager in each branch. It was this employee's responsibility to ensure that each retail customer who entered the branch was guided to a service employee or, alternatively, to a technological interface, such as an ATM, to receive the appro- priate level of service. The redesign also required a large degree of innovation in two further areas: the information system and the telephone call center. Changes in the information system were designed to help relocate and standardize a large number of routine types of account inquiries and transactions and to give National employees a fuller picture of each customer' s financial position and potential. This more com- plete picture of the customer's portfolio was thought to enhance sales efforts, enabling service representatives to suggest a fit between customers and services, and to refer the customers to areas in the bank with expertise in a particular prod- uct. The retail bank branch would be turned into a sales-focused financial store. Challenges in the IT area were heightened by the legacies of existing tech- nologies and the requirement that customer service continue to be provided accu- rately and without interruption; customers are not patient with errors or delayed access to their own money. Over time, a large number of systems, laid one on top of the next, had accumulated in the bank. Further, the redesign had both the advantages and disadvantages of being introduced on the heels of a number of earlier, more piecemeal technological and sales initiatives aimed at the same goals. Both the marketing and IT functions had been continuously seeking to improve National's capabilities in these areas. Support for these initiatives, and their success, had been uneven across the various geographic areas. The Market- ing and IT departments had also worked with a number of other outside vendors. It was not immediately obvious whether the more systematic redesign should complement or substitute for these earlier, more incremental changes in systems or whether these vendors would, or should, have a role in the redesign. Over time, however, these consultants and vendors came under increasing pressure to coordinate their efforts with those of the implementation team, and those who were unsuccessful in doing so were replaced. The importance of the telephone call center raised a new set of challenges. National had lagged a number of its competitors in the sophistication of its tele- phone banking system; yet, through the redesign, it hoped to make telephone banking and eventually, PC or home-banking, cornerstones of its delivery sys- tem. Branch redesign, therefore, also required the construction of new call cen- ters, staffing them as the customers began to be directed toward them, and devel- oping an organizational structure not simply to run the call centers but to manage
98 U.S. INDUSTRYIN2000 the relationship between the call centers and the branches. Even more consult- ants and vendors were required here. The delineation between the new redesign in the branch system and the specialized expertise of the vendors working with telecommunications technology was clearer, so that managing these continuing relationships raised fewer immediate problems than in the case of the branch- based vendors. However, and more recently, as implementation has continued, new challenges have emerged. The increasing importance of the telephone cen- ters has increased the pressures on the call centers for accurate and effective service, even as the call centers struggle with much more basic issues around staffing and the physical implementation of the telecommunications systems. Changes in the physical layout of the branches, in information systems, and in the design of key business processes therefore attracted the attention of the implementation team from the beginning of the innovation process. As plan- ning for the implementation of the pilot redesign proceeded, however, it became increasingly obvious to many on the implementation team that the true anchor for the set of innovations was none of these factors. Most critically, the innova- tions relied upon significant changes in key jobs in the branch systems, on the human resource practices that supported these jobs, and on employees' reactions to these changes. In order to reinforce the idea of standardization across the branch system, and to focus efforts toward sales and efficient delivery of services more clearly, the implementation team recommended that the redesign eliminate the position of local branch manager. In each branch, a customer-relations manager would coordinate customer service efforts, but this person would not have direct author- ity over the tellers and platform employees in branches. Rather, branch employ- ees would report to supervisors by area: customer-relations employees, branch- sales specialists, and tellers each would be assigned to remote leaders. On the platform, a variety of specialized customer service and sales positions were to be consolidated into a position that was eventually titled "Financial Specialist." Local areas were also to be staffed with a few roving financial consultants who did not have specific branch assignments. Only the tellers were to remain rela- tively unscathed by the proposed changes. With this design, the pilot was implemented in two small local markets. Most of the hundreds of administrative and servicing processes were removed from the branch. Telephones no longer rang in the branches. The financial specialists were freed to concentrate on sales activities and found themselves with time avail- able to pursue sales opportunities prospectively rather than simply reacting to walk-in traffic. Most customers responded to the innovation positively, quickly migrating to the new technologies with few problems. The active roles played by the customer-relations managers, many of whom were former branch managers, helped this migration along. The pilot implementation also revealed a number of problems in the design. First, employees and customers in a few of the most rural branch locations met
RETAIL BANKING 199 the redesigned branch with great skepticism. After a period of wrestling with modifications to the design, and considering the benefits associated with the implementation of a single, standardized form of service delivery, the implemen- tation team agreed to abandon the idea of a single best design. National Bank acknowledged that the characteristics of rural markets differed fundamentally from urban and suburban locations. Rural customers, and the way they expected banks and their employees to provide service, were not likely to be served effec- tively by the redesigned branch. A new task force was commissioned to explore this problem and to come up with a design that gained some of the efficiencies associated with standardization and re-engineering for rural branches while ac- knowledging the key differences. A second critical problem was the slow implementation of new technology. Many of the new features of the technology needed to support the new design simply were not ready or did not work as promised. The implementation team, finding it necessary to push forward and being uncertain about when these fea- tures would be ready, moved ahead with the new design anyway once the team was assured that there would not be critical gaps or stoppages in the provision of services. Basic services were satisfactory. The remaining problems related chiefly to ease of use, performance measurement software, and databases and other systems that were intended to provide more support for sales. Third, while most customers adjusted to the new arrangements quickly, and the new processes that were accompanied by supportive technology worked effec- tively, turning the retail bank branch into a sales-focused financial store proved more difficult. Financial specialists found it difficult to move from the idea of reacting to the sales opportunities that routine servicing occasionally provided to the more pro-active role that the redesign called for. Some even claimed that the redesign was responsible for decreased sales as a result of the streamlining. The implementation team wondered in turn how much of this difficulty could be attrib- uted to the design and how much to skills deficits among the financial specialists. A fourth problem was the difficulty in implementing human resource prac- tices necessary to support the new organization. The deficit in skills raised fur- ther issues. For example, training was critical to the success of the implementa- tion, yet the organization had little time to spend in developing the skills critical to the pilot's success. Further, it had been clear that the selection process for new employees would have to be adjusted to seek employees who were more likely to be effective sales agents, but the initial difficulties with the design made this even more imperative. And while incentive compensation systems were also changed to reflect the new goals of the redesign, these were experimental and required considerable fine-tuning. Perhaps most important, however, was that the new jobs had effectively destroyed career ladders in the pilot branches. No longer could tellers easily move to platform positions; these positions were now ex- pected to require an entirely different skill set and, for new applicants, usually a college degree. The financial specialists, who in most cases had been platform
200 U.S. INDUSTRYIN2000 employees, could no longer expect to be promoted to branch management posi- tions: these positions had been abolished and many of the branch managers be- came customer-relations managers. In each functional area, the hierarchy was flattened. While this yielded efficiency gains, it left employees quite uncertain about their future in the organization. The implementation team spent much of its time with the nuts and bolts of the new design. Technological and process-related problems with implementa- tion, and the challenges associated with performance measurement, consumed the team's attention. However, the human resource problems raised serious con- cerns for the longer-range success of the redesign. Employee confusion and skep- ticism over the new design was emerging as an impediment to the success of the innovation and this was occurring in an environment designed to soft-pedal such concerns. Because the team was concerned about the effectiveness of the technological, process, and architectural changes, they decided that the redesign would not be accompanied by any layoffs in the pilot branches. They also knew that, to achieve the eventual efficiencies they expected, some downsizing of the retail bank would be necessary. They did not expect that natural attrition, even in the relatively high-turnover retail bank, would yield the cuts in jobs that they hoped for. The team realized that, in future implementation, the insecurity gener- ated by the job changes would be intensified by the layoffs that would accompany these changes. Despite these problems, the redesign, with some modifications, moved for- ward. A second pilot redesign was implemented in urban and suburban markets in a geographic area distinct from the earlier pilot. More attention was paid to training and selection for the new positions; again, outside consultants were re- lied upon, this time to help identify employees with appropriate skills and to develop those skills. Some of the technological gaps and challenges had been addressed, yet some remained, yielding a new set of complications in the specif- ics of implementation. And the second pilot revealed a new set of problems. In this local area, the situation in the branches before the change differed consider- ably from those in the first set of pilots. In particular, these branches had already been sharply focused on sales opportunities, a reflection of the bank's strategy in this geographic area. While disruption of the status quo in the first set of pilots had been considered to be positive, the benefits of this disruption in the second group which was already moving toward a system of sales-focused branches- were less clear to local managers. Consequently, they were more skeptical about the benefits of redesign and of a standardized model. Local managers consis- tently argued for local adaptation of the model, claiming that they knew best what sorts of processes, technologies, and job structures were likely to be most effec- tive in their area. The implementation team, while sympathetic to these claims, generally re- sisted the pressure to adapt but recognized a further difficulty. To argue that the redesigned model must be strictly adhered to was to admit that no further learning
RETAIL BANKING 201 was to occur as a result of the innovation. They struggled to find ways to differ- entiate between local learning that truly represented a positive improvement to the design concepts and local arguments grounded more in resistance to changing established routines. They also sought principles for making these distinctions as the design was to be rolled out over a much wider area. Currently the team is preparing to implement the new design across the re- mainder of National's retail bank system, with substantial modifications drawn from the lessons learned through the pilots and other issues that have emerged as the process of innovation has continued. Among these challenges are the prob- lems associated with introducing yet another round of innovations in local areas that have already witnessed massive change in recent years as a result of the frantic pace of mergers and acquisitions in the industry. Some of the branches that will be the objects of the redesign will have had three parent banks in the past three years; each change has been accompanied by changes in jobs, processes, systems, and supporting human resource practices. Heaping yet more change onto these locations will be especially difficult. A second challenge facing the implementation team stems from the current decentralized approach to management of the retail bank. While the details of the pilot redesign have not been formally disseminated across the various geographic areas, word that the bank of the future is soon to arrive has traveled widely. Some of the members of the implementation team have returned to management posi- tions in their local areas. Smart local managers have already begun to identify the trends that the implementation team was charged with addressing and have begun to address these challenges locally with their own changes and strategies. Thus the implementation team will be trying to innovate not in a static or standard set of channels but in a wide array of varied and dynamic conditions: in short, against moving targets. Already some local managers have explicitly expressed a desire to get ahead of the game by proceeding with implementation of the features of the pilot redesigns they find most attractive. Left unanswered is how and whether the implementation team will be able to implement other features or how they will reconcile differences in the preemptive local redesigns with their own plan. Appropriately configuring human resource practices to support innovative systems and process changes raises further, significant challenges. On the one hand, it is clear that simply changing job design and pay systems, and coupling these with other technological and system changes, will be insufficient. Attention must also be given to employee selection and promotion systems, training pro- grams, appraisal systems, the use of flexible scheduling, and the bank's overall approach to employee involvement. However, contemplating such sweeping change severely taxes the organization. While piecemeal change in the human resource system is unlikely to yield the results desired, more comprehensive change raises significantly more challenges in implementation. At National, the hope is that investment in the redesign will improve several areas of performance simul- taneously sales effectiveness, productivity, and the quality of customers' rela
202 U.S. INDUSTRYIN2000 tionships with the bank. In practice, this has proven difficult. The early, piloted version of the redesign was effective at serving customers efficiently: the bank streamlined processes and introduced new technological options. However, the effect of the redesign on sales performance and on the overall depth and quality of the customer relationship is not as clearly positive. In fact, some of the streamlin- ing designed to supplement or improve employee-customer interaction may be replacing this interaction. This may mean missed sales opportunities and fewer chances for bank representatives to assess and attempt to meet customers' needs. Because much of the change is held to be a necessary response to continuing competitive pressures, it is unlikely that the redesign will actually be evaluated in strict cost-benefit terms. Such an evaluation of these innovations, their costs and benefits, will require a longitudinal, sustained, consistent effort by the bank, even as members of the implementation team begin to rotate to other positions within the bank. It will also be difficult to decouple the effects of the redesign from other major changes in marketing and product offerings and from the results of continuing merger and acquisition activity. Should the design prove successful, this itself will raise sequential challenges for National, which must further innovate to deliver on the promises raised by successful change. To the extent that customers are convinced to migrate to alternative, more efficient delivery channels, the bank must continue to develop its ability to manage those channels effectively. Such channels, particularly tele- phone and PC banking, are not only more technology-intensive but also raise new sets of organizational and human resource problems. As the use of such channels grows, and as their range of functions increases, questions over appropriate staff- ing, training, performance measurement, and reporting structures multiply. Inno- vation, both organizational and technological, may actually have to intensify as a result of the success of prior changes. Where's R&D? The Process of Innovation in a Bank The two examples given above highlight the complex organizational design issues involved in the innovation processes in retail banking. Simply put, most retail banks do not have something called an R&D group. If they do, these groups play an important, but small, role in the overall innovation practices of the orga- nizations. Marketing, business units, information technology, and a complex web of information technology suppliers and consultants drive the innovation pro- cesses in banking. Consider the case of National Bank, where there was no division devoted to thinking about or implementing innovation, no "research and development" or similar functional structure. Rather, pressure for innovation built incrementally as a result of numerous smaller initiatives by marketing, by those responsible for managing technological systems, and by line managers. Each area felt competi- tive pressure and began to develop responses. At National Bank, these responses
RETAIL BANKING 203 were eventually, to some extent, collected and channeled through the implemen- tation team, although they also maintained some momentum of their own. At National Bank, translating this pressure to innovate into actual techno- logical and organizational changes was greatly facilitated by the continuing pres- ence of consultants and of suppliers of technology. Indeed, one way to under- stand at least part of the role of consultants is that they function as suppliers of the organizational technology required to leverage the potential gains from innova- tions in computing and telecommunications systems. While the organization con- tinues to develop its capacity to learn and innovate, it explicitly recognizes that it has considerable distance to travel in order to exercise this capacity more inde- pendently. One further lesson we take from National in the midst of this redesign is that changes in IT, and in technological capabilities can spark the desire for system- wide innovation and even shape its particular form. With the enthusiastic promo- tion of consultants and outside vendors, technology is perceived by retail banks to be a catalyst for change across the organization. Yet even where this technology is over-sold, poorly understood, or fails to deliver on its promises, the process of innovation may take on its own momentum. In the case of PC banking, such organizational changes are heightened by the presence of external suppliers of technology, consumer access, and fulfillment services. As banks continue to grapple with the variety of choices for electronic delivery, new organizational forms and entities are sure to emerge. As an ex- ample, the Bank of Montreal recently created a direct bank called mbanx,7 whose purpose is to be a non-branch-based deliverer of financial services that will di- rectly compete with the existing Bank of Montreal delivery and sales organiza- tion. Such developments of new organizational systems for non-physical deliv- ery are sure to accelerate in the next decade. WHAT CONSTITUTES EFFICIENT INNOVATION? l o produce innovation in the banking industry, complex organizational struc- tures are needed. Given this context, which banks are efficient at such innova- tion? To address this issue, Prasad and Harker (1997) consider the overall impact of information technology on productivity in the retail banking industry in the United States. Using a Cobb-Douglas production function, Prasad and Harker estimate the following equation with a combination of publicly available and proprietary data: where Q = output of the firm, C = IT capital investment, Q= e~OC~K~2S~3LA4 7For details on mbanx, see the following Web address: http://www.mbanx.com/. (1)
204 U.S. INDUSTRYIN2000 K= non-IT capital investment, S = IT labor expenses, L = non-IT labor expenses, and pi, p2, p3, and p4 are the associated output elasticities. Using this function, the following three hypotheses were tested: · IT investment contribution to output is positive (that is, the gross mar- ginal product is positive); . IT investment contribution to output is positive after deducting deprecia- tion and labor expenses (that is, the net marginal product is positive); and · IT investment makes no contribution to the firm's profits or stock market value. Just what constitutes a bank's output is a subject of some discussion. Sum- ming up the problem Benston et al. (1982) posit that, "output should be measured in terms of what banks do that cause operating expenses to be incurred." Various studies of productivity have taken various approaches to this question, and they may be classified into three broad categories the assets approach, the user-cost approach, and the value-added approach (Berger and Humphrey, 1992~. Prasad and Harker look at a wide variety of output measures, both financial and cus- tomer satisfaction. The most meaningful results from this analysis arise when total loan + deposits is used as the output of the institution; these results are summarized in Table 5. From this table, it can be seen that the elasticities or coefficients associated with IT capital and labor are positive. However, the low significance associated with the IT capital coefficient implies that there is a high probability (0.93) that the elasticity of IT capital is zero. Thus, there is not sufficient evidence to sup- port the hypothesis that IT capital produces positive returns in productivity. It is interesting to note that the elasticity of non-IT capital is at best zero (being not significantly different from zero), implying that IT capital investment is rela TABLE 5 Results of the Estimation of Equation 1 When Output = (Total Loans + Total Deposits) Parameter Coefficient Standard Error t-statistic t-statistic significance (%) Ratio to Marginal output product 0.000452 0.0006 0.00428 0.01475 IT capital 0.00116 IT labor 0.25989 Non-IT capital -0.02071 Non-IT labor 0.53244 Note: R2 = 41% (OLS); 99% (2-step WLS) Source: Prasad and Harker (1997). 0.013 0.089 0.031 8.34 0.026 -0.79 0.059 8.95 7 100 100 2.56 449.75 -4.84 36.10
RETAIL BANKING 205 lively better than investment in non-IT capital. However, since the marginal product of IT labor is $449.75, it can be concluded that IT labor is associated with a high increase in the output of the bank. Since the first hypothesis cannot be supported for IT capital, the discussion of the stronger hypotheses, the second in the list, is restricted to the IT labor results. First, it can be seen that the marginal product for IT labor is very high. Since IT labor is a flow variable, then every dollar of IT labor costs a dollar. In view of this, the excess returns from IT labor can be computed to be $~449.75 - 1), or $448.75. Thus, this hypothesis cannot be rejected for IT labor. For the last hypothesis, one has p3 - (IT labor expenses / non-IT labor expenses ~ x D4 = 0.2390 > 0. Thus, there is support for the claim that investment in IT labor makes a positive economic contribution. As far as capital expenses are concerned, it can be seen that the marginal product of non-IT capital is negative. Further, given the standard errors of the estimation, it is asserted that IT capital is more likely to yield either slightly positive or no benefits, whereas non-IT capital will most probably have a nega- tive effect, decreasing productivity. More formally, p~ - (IT labor expenses / non-IT labor expenses ~ x D2 = 0.0034 > 0. Given the significance associated with the IT capital estimate, however, the last hypothesis failed to be rejected. Thus, these results show no strong evidence of IT capital making a positive contribution to output. This result is significantly different from previous studies in the manufacturing sector (Lichtenberg, 1995; Brynjolfsson and Hilt, 1996), and seems to be more in conformity with those obtained by Parsons et al. (1993), the only formal study on IT in banking to date. While Parsons et al. report a slightly positive contribution to IT investment, this analysis demonstrates zero or slightly negative contributions. IT labor presents a very different picture than does IT capital. IT labor con- tributes significantly to output; its marginal product is at least 10 times as much as that of non-IT labor. Rather than make the simplistic conclusion from this that a single IT person is equivalent to 10 non-IT persons, it is better perhaps to specu- late that this may simply reflect the fact that there is a significant difference be- tween the types of personnel involved in IT and non-IT functions. It is more interesting to compare the marginal product of IT capital versus IT labor. It is striking that while IT labor contributes significantly to productivity increases, IT capital does not. Thus, these results state that while the banks in our study may have over-invested in IT capital, there is significant benefit in hiring and retain- ing IT labor. This result and interpretation is consistent with the idea that aligning capital, rather than throwing technology at problems, is what affects efficiency. IT per
206 U.S. INDUSTRYIN2000 sonnet are likely to be much more effective at ensuring that the implementation of technology does what it is meant to do. The general point is that the manage- ment of IT has profound effects on efficiency. Banks that are able to manage their IT effectively are likely to be efficient. These results are consistent with our fieldwork experiences. They are also consistent with the fact that today's high demand for IT personnel is unprecedented in U.S. labor history. Figures from the U.S. Bureau of Labor Statistics show that while the overall job growth in the U.S. economy was 1.6 percent between 1987 and 1994, software employment grew in these years at 9.6 percent every year and jumped to 11.5 percent in 1995. The prediction is that, over the next decade, software employment will grow 6.4 per- cent every year (Rebello, 1996~. The problems are actually likely to be more subtle than our measures sug- gest. For example, IT personnel, while evidently valuable, may not be equally valuable. The point was driven home to us in a series of interviews at a major New York bank. A Senior Vice President there lamented the fact that, "The skills mix of the IT staff doesn't match the current strategy of the bank," and said that he "didn't know what to do about it." At the same bank, the Vice President in charge of IT claimed, "Our current IT training isn't working. We never spend anywhere near our training budget." IT labor is in very short supply, and issues as basic as re-skilling the workforce cannot be addressed given the lack of suffi- cient IT labor in banking. Other researchers have observed this dependence and under-investment in human capital in technologically intensive environments. To quote Gunn' s (1987) work in manufacturing, "Time and again, the major impediment to [technologi- cal] implementation . . . is people: their lack of knowledge, their resistance to change, or simply their lack of ability to quickly absorb the vast multitude of new technologies, philosophies, ideas, and practices that have come about in manu- facturing over the last five to ten years." Another observation about the transi- tions firms need to make to gain from technology comes from Reich (1984), again in the manufacturing context: "The transition also requires a massive change in the skills of American labor, requiring investments in human capital beyond the capital of any individual firm." The evidence also suggests that the effects of managing IT are being felt more broadly in a retail bank. Consider the inclusive model for managing branches. In this model, IT and process redesign (so-called re-engineering) com- bine to remove as many basic servicing tasks as possible from employees. These tasks simple inquiries, transactions, and movement of funds can be automated or turned over to customers. Re-engineering frees employees to concentrate more effort on activities that have potentially higher added value customized transac- tions and the provision of financial advice coupled with sales efforts. Second, IT gives each employee a full picture of each customer's financial position and potential. This enhances sales efforts, enabling tellers and customer service rep- resentatives to suggest a fit between customers and services and to refer the
RETAIL BANKING 207 customers to employee teammates with particular expertise in a product, as necessary. Challenges under the segmented model are less acute, yet still present. In this model, technology is used to simplify the majority of the jobs to make them easier to learn and, therefore, to make turnover less costly. Only those bank employees in high value-added, personal banking jobs have access to the broad range of information that might be useful in generating sales leads and opportunities. In order for either model to function effectively, those responsible for de- signing IT must understand not only the purposes of the technology but the capa- bilities and propensities of the workforce and the likely effects of different choices in technology on employee and customer behavior. Further, IT staff must be able to assess the likely effects of different configurations of technologies and em- ployment systems if they are to be able to contribute to strategic decisions around the deployment of IT. Thus, our results are very consistent with Osterman's (1986) conclusion that ". . . as IT capital prices fall, production becomes increasingly information-worker intensive." Our results seem to confirm this; increasing IT investments in banks requires a substantial investment in IT labor. Further, IT labor is the most profit- able of all four types of investment IT and non-IT capital and labor available to the bank. Accordingly, the biggest challenge facing banks with respect to efficient and effective innovation lies in the management of the "New Age Indus- trial Engineers" that must combine technological knowledge with process design in order to create the delivery systems of the future. BANKING INNOVATIONS: LESSONS FOR THE STUDY OF SERVICES Our study of banking innovation leads us to reconsider the basic model of innovation in the standard textbooks and readings in the field.8 While the basic steps of the innovation process, such as those outlined by Marquis (1969), remain the same, the change arises in the combination of actors who perform these steps. The standard view is that R&D, operations, and marketing combine in a complex web of interactions to generate innovation (Figure 10~. However, as we have seen from our previous discussion, vendors that supply outsourced services and technology play a vital role in this innovation process.9 Even more important in the development of innovations is the role of the "sys- tems integrator" the person or organization that pulls together not only the op ~For example, the collection of readings in Tushman and Moore (1988). 9For a discussion on the strategic role of firms that supply outsourcing services, see, for example, Jonash (1996), Chesbrough and Teece (1996), and Rubenstein (1994). For the particular case of financial services, see Drew (1995).
208 U.S. INDUSTRYIN2000 FIGURE 10 Basic relationship in innovation processes. Source: Adapted from Galbraith (1982). orations, IT, and marketing functions for a single innovation but also manages the portfolio of innovations in the organization. At National Bank, this systems inte- gration role is played by an in-house re-engineering team in conjunction with their external consultant (see Figure 1 1~. Ultimately, it is this systems integration function that will make or break innovation efforts. Jonash (1996) argues that the systems integration function belongs in the hands of the chief technology officer who will coordinate the ef- forts of internal and external innovation efforts for the benefit of the organization. The discussion in the previous section about the critical role of the IT organiza- tion in the overall efficiency of the banks tends to support this view. External Vendors & Consultants 1 ~ - - R&D )< Systems In integrators - - External Vendors & Consultants Marketing ) FIGURE 11 A new model for innovation involving expanded relations.
RETAIL BANKING 209 The role of the systems integrator is crucial for the future of retail banking. Frei et al. (1997), in summarizing their various analyses of retail banking effi- ciency based on the dataset described in the Appendix of this chapter, paint a picture of what makes an effective bank. They conclude that there is simply no single set of management practices, capital investments, and strategies that lead to success. Rather, it appears that the "devil" is truly in the details. The align- ment of technology, human resources management, and capital investments with an appropriate production "technology" appears to be the key to efficiency in this industry. To achieve this alignment, banks need to invest in a cadre of "organiza- tional architects" that are capable of integrating these varied pieces together to form a coherent structure. In fact, several leading financial services firms have realized the need for such talents and are investing heavily in senior managers from outside the industry most notably, from manufacturing enterprises to drive this alignment of technology, human resources management, and strategy. The challenge, therefore, is not to undertake any one set of practices but rather to develop senior management talent that is capable of this alignment of practices. This is especially important because the future direction of the industry is subject to a tremendous degree of uncertainty. This uncertainty was revealed in the strategy-related data we collected as part of this study. We found that most banks simply could not articulate a consistent and coherent strategy for the fu- ture.~° In numerous visits with the banks that were a part of the study, we would feed back the data that bank officials had given to us in order to check their validity. When we came to our survey' s strategy-related questions, someone in the bank, usually at a senior management level, would state something along these lines: "This is wrong. This CAN'T be our strategy!" We would tell them who provided these data (always another senior manager), and we would become embroiled in a debate over defining the strategy of the bank! This confusion reflects the tension between making investments to perfect today's strategy versus investing in a portfolio of alternative strategies for the future. This tension is both quite typical and quite real in the banking industry. Given the inability to control the use of the varied distribution channels, includ- ing ATMs and branches, banks are either investing in all channels simultaneously or undertaking fairly radical changes in their service offerings to deal with this proliferation of services. Thus, bank managers face a crucial decision as to miss- ing the "correct" strategy for the future versus living with misaligned systems that they know to be inefficient. Given this uncertainty, the removal of inefficient firms may take quite a long time. Furthermore, if we are correct in our assessment that a major cause of inefficiency in the industry is the misalignment of management practices, a major cause of persistent inefficiency in the banking industry may be the necessity for integrated financial services organizations to "hedge their bets" on the future. Thee Hunter (1996), in the context of human resources.
210 U.S. INDUSTRYIN2000 Clearly, alignment would be simpler and more rapid in an industry made up of many "niche" players, each focusing on a likely future scenano. Such movement to disintegrate financial services is already under way in most banking organiza- tions; business units such as credit cards and trust divisions are now being run as completely separate operations. The bottom line of this analysis is that service industries such as banking must develop a new generation of management talent to play the role of systems architect one who can blend technical knowledge with complex organizational design issues to drive innovation through their firms. REFERENCES Akhavein, J.D., A.N. Berger and D.B. Humphrey. (1997). "The effects of megamergers on efficiency and prices: evidence from a bank profit function," Review of Industrial Organization 12:95- 139. Benston, G.J., G.A. Hanweck, and D.B. Humphrey. (1982). "Scale economies in banking: a restruc- turing and reassessment," Journal of Money, Credit and Banking 14:435-450. Berger, A.N., D. Hancock, and D.B. Humphrey. (1993). "Bank efficiency derived from the profit function," Journal of Banking and Finance 17:317-348. Berger, A.N., and D.B. Humphrey. (1992). "Measurement and efficiency issues in commercial bank- ing," in Output Measurement in the Services Sector: National Bureau of Economic Research Studies in Income and Wealth, Z. Griliches ea., Chicago: University of Chicago Press. Berger, A.N., W.C. Hunter, and S.G. Timme. (1993). "The efficiency of financial institutions: a review and preview of research past, present and future," Journal of Banking and Finance 17:221-250. Berger, A.N., A.K. Kashyap, and J.M. Scalise. (1995). "The transformation of the U.S. banking in- dustry: what a long, strange trip it's been," Brookings Papers on Economic Activity 2:55-218. Brynjolfsson, E. and L. Hitt. (1993). "Is information systems spending productive? New evidence and new results," Working Paper, Coordination Laboratory, MIT Cambridge, MA. Brynjolfsson, E., and L. Hitt. (1996). "Paradox lost? Firm-level evidence on the returns to information systems spending," Management Science 42:541-558. Cates, D.C. (1991). "Can bank mergers build shareholder value?" Journal of Bank Accounting and Finance 6-7. Chesbrough, H.W., and D.J. Teece (1996). "When is virtual virtuous? Organizing for innovation," Harvard Business Review 74(January-February):65-71. Council on Financial Competition. (1996). Letter from the Future: Beyond the Branch-Based Fran- chise, Washington, DC: The Advisory Board Company. Drew, S.A.W. (1995). "Accelerating innovation in financial services," Long Range Planning 28:11 21. Ernst and Young. (1996). Creating the Value Network 1996, New York. Frei, F.X. (1996). The Role of Process Designs in Efficiency Analysis: An Empirical Investigation of the Retail Banking, unpublished Ph.D. dissertation, The Wharton School, University of Penn- sylvania. Philadelphia. Frei, F.X., P.T. Harker, and L.W. Hunter. (1997). "Inside the black box: what makes a bank effi- cient?" Working Paper 97-20, Wharton Financial Institutions Center, The Wharton School, University of Pennsylvania, Philadelphia; also available at http://wrdsenet.wharton.upenn .edu/fic/wfic/papers.html Frei, F.X., and R. Kalakota. (1997). "Frontiers of Online Financial Services," in Banking and Finance on the Internet M. J. Cronin ea., New York: Van Nostrand Reinhold Press.
RETAIL BANKING 211 Fried, H.O., C.A.K. Lovell, and P.V. van Eeckaut. (1993). "Evaluating the performance of U.S. credit unions," Journal of Banking and Finance 17:251-266. Galbraith, J.R. (1982). "Designing the innovating organization," Organizational Dynamics (Win- ter):3-24. Griliches, Z. (1992). Output Measurement in the Services Sector: National Bureau of Economic Re- search Studies in Income and Wealth Chicago: University of Chicago Press. Gunn, T.G. (1987). Manufacturing for Competitive Advantage, Cambridge, MA: Ballinger. Herring, R.J., and A.M. Santomero. (1991). "The role of the financial sector in economic perfor- mance," Study Prepared for the Kingdom of Sweden's Productivity Commission, Stockholm. Hitt, L., and E. Brynjolfsson. (1996). "Productivity, business profitability, and consumer surplus: three different measures of information technology value," MIS Quarterly (June):121-142. Huber, G.P., and D.J. Power. (1985). "Retrospective reports of strategic-level managers: guidelines for increasing their accuracy," Strategic Management Journal 6:171-180. Hunter, L.W. (1996). "When fit doesn't happen: The limits of business strategy as an explanation for variety in human resource management practices," presented at the Academy of Management Annual Meeting, Cincinnati, Ohio, August 1996. Jonash, R.S. (1996). "Strategic leveraging making outsourcing work for you," Research-Technology Management 39:19-25. Kennickell, A.B., and M.L. Kwast. (1997). "Who uses electronic banking? Results from the 1995 survey of consumer finances," Working Paper, Division of Research and Statistics, Board of Governors of the Federal Reserve System, Washington, DC. Leibenstein, H. (1966). "Allocative efficiency verses 'X-inefficiency," American Economic Review 56:392-415. Leibenstein, H. (1980). "X-efficiency, intrafirm behavior, and growth," in Lagging Productivity Growth, S. Maital and N. Meltz eds., Cambridge, MA: Ballinger Publishing. Lichtenberg, F.R. (1995). "The output contributions of computer equipment and personnel: a firm- level analysis," Economics of Innovation and New Technology 3. Marquis, D.G. (1969). "The anatomy of successful innovations," Innovation (November). "Mutual Fund Review," Wall Street Journal, April 1996. National Research Council. (1994). Information Technology in the Service Society, Washington, DC: National Academy Press. Oliver Wyman and Company. (1997). Personnel communication. Osterman, P. (1986). "The impact of computers on the employment of clerks and managers," Indus- trial and Labor Relations Review 39: 175-86. Parsons, D., C.C. Gotlieb, and M. Denny (1993). "Productivity and computers in Canadian banking," in Productivity Issues in Services at the Micro Level. Z. Griliches and J. Mairesse eds., Boston: Kluwer Academic. Peristiani, S. (1997). "Do mergers improve X-efficiency and scale efficiency of U.S. banks? Evidence from the 1980s," Journal of Money, Credit, and Banking 29:326-337. Perrin, T. (1996) "Letter from the Future: Beyond the Branch-Based Franchise," Council on Financial Competition. Washington, DC: The Advisory Board Company. Prasad, B., and P.T. Harker. (1997). "Examining the contribution of information technology toward productivity and profitability in U.S. retail banking," Working Paper 97-09, Financial Institu- tions Center, The Wharton School, University of Pennsylvania. Philadelphia; available at http://wrdsenet.wharton.upenn.edu/fic/wfic/papers.html Pine, B.J. (1993). Mass Customization: The New Frontier in Business Competition. Boston, Harvard Business School Press. Rebello, K. (1996). "We humbly beg you to take this job. Please," Business Week (June). Reich, R.B. (1984). The Next American Frontier. New York: Penguin Books. Rhoades, S.A. (1993). "Efficiency effects of horizontal (in-market) bank mergers," Journal of Bank- ing and Finance 17:411-422.
212 U.S. INDUSTRYIN2000 Roach, S.A. (1993). Making Technology Work. Economic Research Unit, Morgan Stanley & Co., New York. Rubenstein, A.H. (1994). "Trends in technology management revisited," IEEE Transactions on Engi- neering Management 41:335-341. Shaffer, S. (1993). "Can mergers improve bank efficiency?" Journal of Banking and Finance 17:423- 436. Singh, H., and M. Zollo (1997). "Learning to acquire: knowledge accumulation mechanisms and the evolution of post-acquisition integration strategies," Working Paper 97-1OB, Financial Institu- tions Center, The Wharton School, University of Pennsylvania. Philadelphia; available at http://wrdsenet.wharton.upenn.edu/fic/wfic/papers.html. Tushman, M.L., and W.L. Moore, eds. (1988). Readings in the Management of Innovation, 2nd Edi- tion. New York: Harper Business. U.S. Bureau of Labor Statistics. (1992). Employment and Earnings Report, March.
RETAIL BANKING 213 APPENDIX: STRUCTURE OF THE WHARTON/SLOAN RETAIL BANKING STUDY This paper is partially a result of the work undertaken by the retail banking study at the Wharton Financial Institutions Center. The retail banking study is an interdisciplinary research effort aimed at furthering the understanding of com- petitiveness in the industry, where competitiveness means not simply firm perfor- mance but the relationship between industry trends and the experiences of the retail banking labor force. In the exploratory first phase of a study of the U.S. retail banking industry during summer 1993 through fall 1994, a research team conducted open-ended and structured interviews with industry informants and shared its impressions with these informants at a number of conferences. The team interviewed top executives, line managers in retail banking, human resource managers, executives responsible for the implementation of information technology, retail bank employees, and industry consultants. The first phase fea- tured site visits to thirteen U.S. retail bank headquarters and interviews with nu- merous other managers and employees in remote and off-site locations. The interviews began with very general questions, and the questions increased in specificity as the research progressed. In this phase of the study, the team col- lected data through the use of two waves of structured questionnaires in seven retail banks. The team's analysis of the data in these questionnaires was then presented to management teams in six of the seven banks and was used as the basis for the second phase, a large-sample survey. This detailed survey addressed technology, work practices, organizational strategy, and performance in 135 U.S. retail banks, chosen to yield the broadest coverage of trends in human resources, technology, and competitiveness in the industry. The survey focused on the largest banks in the country and was not intended as a random sample of all U.S. banks. In the end, the approach gained the participation of banks holding over three-quarters of the total assets in the industry in 1994. The process began by compiling a list of the 400 largest bank holding companies (BHCs) in the United States at the beginning of 1994. Merger activity, and the fact that a number of BHCs had no retail banking organization (defined as an entity that provides financial services to individual consumers), reduced the possible sample to 335 BHCs. Participation in the study was confi- dential but not anonymous, enabling the team to match survey data with data from publicly available sources. Participation in the study required substantial time and effort on the part of organizations. Therefore, commitment to participation was sought by approach- ing the 70 largest U.S. BHCs directly; in the second half of 1994, we requested the participation of one retail banking entity from each BHC. Fifty-seven BHCs agreed to participate. Of these, seven BHCs engaged the participation of two or more retail banks in the BHC, giving us a total of 64 participating retail banks.
214 U.S. INDUSTRYIN2000 Multiple questionnaires were delivered to each organization in this sample. Ques- tionnaires ranged from 10 to 30 pages and were designed to target the "most informed respondent" (Huber and Power, 1985) in the bank in a number of areas, including business strategy, technology, human resource management and opera- tions, and the design of business processes. The team made a telephone help line available to respondents who were unsure of the meaning of particular questions. Questionnaires were delivered to four top managers: the head of the retail bank, the top finance officer, the top marketing officer, and the top manager responsible for technology and information systems. These banks received questionnaires for one manager of a bank telephone center and for one branch manager and one customer service representative (CSR) in the bank's head office branch, defined as the branch closest to the bank's headquarters. In addition, an on-site researcher gathered data about all business process flows in the head-office branch. Identi- cal questionnaires were mailed to five more branch managers; the instructions to the bank were to choose the sample branches so that, if possible, data were re- ceived from two rural, two urban, and two suburban branches. Questionnaires were also mailed to CSRs in those branches. In these questionnaires, the CSRs themselves mapped processes associated with home equity loans, checking ac- counts, certificates of deposit, mutual fund accounts, and small business loans. In order to facilitate the creation of process maps via the mailed survey, a worksheet was developed for the CSRs to fill out. These worksheets, a sample of which is shown by Frei (1996), list the majority of potential steps required in the process so that the CSR need only indicate the order of the step, the person re- sponsible for its execution, the type of technology involved, and the amount of time the step takes. Adequate space was provided for the addition of steps unique to an institution. In late 1994, survey questionnaires were mailed to top executives of the 265 next largest BHCs and followed with a telephone call requesting the participation of one of their retail banking organizations. Sixty-four of these BHCs agreed to participate in the study, and four of these engaged the participation of two or more retail banks in the BHC, so that a total of 71 retail banks participated in the mailed survey. For this group of banks, the head of the retail bank was surveyed, and many of the questions directed to the other top managers were consolidated into this survey. Prior interviews had suggested that, for banks of this size, the head of retail was able to answer this broader set of questions accurately. For this sample, questionnaires were mailed to one telephone center manager, one branch manager, and one CSR in the head office branch. The telephone help line was also available to respondents in this sample. All together, the entire survey of retail banking covers 121 BHCs and 135 banks, which together comprise over three-quarters of the total industry, as mea- sured by asset size. The scope and scale of this survey make it the most compre- hensive survey to date on the retail banking industry.