The financial services industry, in particular the retail banking sector, is expanding its embrace of electronic distribution strategies. Today, for example, almost all retail banks offer on-line inquiry for customer accounts, and most provide the ability to perform balance transfers, withdrawals, and bill payments. Electronic distribution is growing, too. In 1996, PC banking was estimated to reach about 2.5% of banking households and was projected to grow to approximately 19% by 2002. Early work on electronic commerce and the vast majority of practitioner literature point to cost savings yielded by an on-line distribution approach. For example, a study by
In a paper titled “Do Better Customers Utilize Electronic Distribution Channels: The Case of PC Banking,” Wharton’s Lorin M. Hitt and Harvard’s Frances X. Frei challenge the assumption that cost saving represents the primary long-term benefit offered by electronic distribution in the retail banking sector.
For instance, even if the Booz Allen cost-saving figures are accepted, Hitt and Frei argue that on-line banking technology is easily replicated by competitors and is therefore unlikely to create a sustainable competitive advantage. In addition, many of these cost reductions are accompanied by reduced entry barriers and increased market transparency, which further intensify competition. Finally, they say, the substantial investment in infrastructure as well as incremental service and support costs erode much, if not all, of the supposed cost savings.
While Hitt and Frei reject the cost-savings potential, however, they conclude that PC banking does offer at least one significant benefit—it lets institutions retain high-profit customers.
In 1998, Hitt and Frei developed their own data, convincing seven banks with assets ranging from $30 billion to $200 billion to participate in a comprehensive study of IT (information technology) investment practices. The study included a general overview of the process as well as an extensive discussion and data-collection effort of PC banking as an example of a recent IT investment decision. These data included project timelines, initial and ongoing costs, motivation for the investment and measured outcomes.
Some of the data were collected by a formal questionnaire with specific queries, while other data were obtained in response to such open-ended questions as, “What do you perceive the value of PC banking to be?”
A questionnaire was sent to each bank and interviews were then scheduled with the individual or group most knowledgeable about a particular issue. The authors also collected a large sample of customer records from each bank in order to facilitate the examination of the way that customer characteristics, such as demographics and product usage, differ between customers that use electronic banking and those that do not.
These differences were studied by estimating regression models that related customer demographics and PC banking usage to product balances and the number of products they use. Based on the statistics they developed, Hitt and Frei made three initial observations: First, customers who use PC banking are consistently wealthier, more likely to own a home and to be married, and are between two to six years younger than those who do not use PC banking.
Second, PC banking users generally have more banking products, greater asset balances, and offer greater (estimated) profitability.
Third, the differences between PC banking and non-PC banking customers persist across different institutions in the Hitt-Frei study, although there is some variation on the degree of the difference.
Consequently, say the authors, the value of PC banking may lie in its appeal to high value customers, rather than any changes in customer behavior or value brought about by the technology itself. Hitt and Frei observe that even after accounting for demographic differences, account duration and short-run versus long-run profitability, PC banking customers offer more value than those who do not bank by PC.
For example, electronic customers use more products and carry higher asset and liability balances in both a single cross-section and over different stages of the account lifetime. Following their adoption of PC banking, they also tend to acquire products at a slightly greater rate than that suggested by their characteristics.
The fact that the results are significant across multiple banks suggests a general phenomenon, rather than being characteristic of a single PC-banking implementation, according to the authors.
Their study offers valuable guidance for banks that are considering expanding or implementing electronic distribution. Although the cost-savings benefits of PC banking appear to be minimal or non-existent in the medium term, the product could be viewed as a tool for retaining a small but significant segment of high-value accounts.
In essence, electronic banking is a competitive necessity that offers a high degree of value. Additionally, the product is a useful segmentation device that could open the door to targeted cross-sells, price discrimination and other effective strategies.
Interestingly, to the authors’ knowledge, other than providing the PC banking product free as part of a package of services, no bank currently offers different products, prices or promotions to PC banking customers. Thus, most of the potential segmentation effects are not being realized.
They also note that account retention was a motivation in only four of the seven banks studied. And even in these institutions, there was little measurable explicit pursuit of account retention in the product marketing and deployment approaches.
The authors conclude that the PC banking market presents an opportunity for low-cost and targeted entry, particularly for institutions that are not hampered by existing traditional infrastructure. As such, traditional banks are likely to face new and aggressive competition for high-value electronic customers from on-line virtual banks as well as such so-called mega-banks as Citigroup and NationsBank that have national brand recognition but currently have a limited geographic scope.
