Case study: improving branch profitability and service with data envelopment analysis: data envelopment analysis makes it possible to balance branch expenses with service quality.

How do you manage and improve the profitability and service quality of a network of hundreds or thousands of bank branches disbursed over several states and countries? This article describes a very powerful technique that helps improve service performance, data envelopment analysis (DEA). DEA has generated substantial cost savings in many financial institutions, including three of the largest 50 banks in the United States, as well as in brokerage firms, mortgage banks and other service branch networks. While DEA has generated documented benefits since the early 1990s, it is now more accessible than ever with the availability of Microsoft Excel plug-ins.

Cost management is understandably a secondary concern to the questions of survival that have plagued financial institutions since 2007. As the bailout process proceeds, however, cost management to improve profitability will be one path to restoring the shareholder value. This article presents a case history that shows how DEA generated substantial costs savings for a 200-branch bank network.

Assume that you are responsible for 1,500 or more bank branches operating in 10 states in the United States. Would you be interested in knowing that some branches are using more than twice the resources of other branches while offering the same or lower-quality service than the less costly branch? DEA can reveal this information. DEA (1) locates the best-practice branches that are low cost and high quality, (2) identifies the high-cost branches that could reduce operating cost while maintaining or improving quality and (3) specifies the changes that can be made to reduce costs to the best-practice level. DEA is able to compare each one of the 1,500 branches to every other branch and simultaneously consider all the types of resources used and all the types of transactions and services provided by each branch. This ability generates new paths to improve profitability.

DEA detects opportunities for savings that will endure because they result from installing best-practice methods used by the best service providers in the network. This is in striking contrast to savings from desperate across-the-board layoffs and expense cuts. In a 1,500-branch network, DEA identified excess annual operating costs totaling more than $100 million with an individual prescription to achieve savings for each high-cost branch.

Banks with large branch networks along with banks with fewer than 50 branches are not able to identify best-practice low-cost branches and the inefficient high-cost branches that diminish earnings. The challenge of optimizing branch operations is particularly acute in banks where the branch network is distributed over many states and countries and where branches are being acquired.

How can management be unaware that some branches are incurring excess operating costs compared to others? First, the branches may be geographically far apart--as much as 3,000 miles--and bank analytic systems (and politics) do not encourage and sometimes do not even allow such comparisons. While benchmarking with less powerful methods than DEA does occur, generally only branches in the same regions or district or with the same manager are compared. Second, the techniques most widely used provide a sense that the branches are already well run, because (1) techniques to identify best practices are unable to consider and analyze the complex set of branch transactions and (2) the techniques only identify average performance and not the best practice.

Existing techniques such as queuing models and staffing models do not consider the complex mix of services and resources used by a bank branch and are unable to address the more complex services such as accepting loan applications and opening new accounts. Other methods also fail to consider the full range of services provided by the bank branch or other service unit, resulting in a splintered measure of operations and overlooking the synergies of balancing the mix of services and the mix of responsibilities assigned to service providers.