Several Roads Lead to Lower-Cost Data Processing
Banks everywhere are under intense pressure to slash noninterest expenses. Because most bankers view front-office cutbacks as a direct threat to revenue production, back-office functions -- including data processing -- are often the first to face cost-cutting scrutiny.
Seeking to cut processing costs? Here are several cost-cutting approaches to consider in the back room:
* Outsourcing: Perhaps the most important question for a bank is whether it should be processing its own data at all. At many banks, outsourcing -- the decision to let a third party handle processing -- can save as much as 50%.
But outsourcing is not for everyone. Some banks can process as efficiently as outside vendors. Or they may have superior computer systems that give them a strategic advantage over other banks.
If the decision is made to maintain control over data processing, banks must consolidate data centers to reduce the total as much as possible. Each consolidation can reduce costs by 35% or more.
Studies have shown that, for some banks, two is the optimum number of data centers. But a superregional bank recently consolidated into one center and saved $9 million annually.
* Standardization: Many banks continue to run multiple demand-deposit accounts, known as DDAs, along with loan and financial-reporting systems taken on through acquisitions. But standardizing on one set of application systems can eliminate entire programming teams and package licensing fees. For example, one superregional bank recently halved its staff by eliminating redundant systems.
* Telecommunications: Over time, many voice and data communications networks degenerate into a tangle of inefficiency. But a fresh look at network topology can often cut costs by 30% to 40%.
Hundreds of communications lines can be consolidated into a private, backbone network comprising a few high-volume lines that carry both voice and data communications. This backbone connects data centers and operations centers.
After that, branches can be tied into the closest data or operations center via low-cost lines. The end result is a streamlined network that provides more protection against line outages, while reducing cost. The process can be completed within 12 months.
* Software: Many banks have customized systems that are costly to maintain. In many cases, these systems use old technologies that are difficult to modify.
Typically, too, customized systems are poorly documented. Banks using them are forced to rely on programmers, who have the secrets of these systems locked up in their heads.
An alternative: Use software packages maintained by vendors. These packages spread software costs among many users.
Banks pay maintenance fees for using such software and can keep in-house programming staffs to a minimum. Both flexibility and utility are increased.
* Bricks and Mortar: Building and real estate costs for data centers represent about 25% of most data-processing budgets. And historically, most banks have placed programmers in expensive, highly secure data centers.
With advancements in telecommunications, it is no longer necessary for programmers to be located near the computers. In fact, it is preferable to locate them close to users of the systems; this concept can improve the translation of business needs into working systems.
* Reengineering: Banks can avoid expensive development projects by reengineering, or converting obsolete systems into modern technologies.
Under reengineering, code from the current system is put through proprietary -- "black box" -- software, by a third party. The software translates the code into newer programming languages and data-base technologies.
After translation, the updated code is returned to the bank. This can be done at a fraction of the cost of building a new system from scratch.
* Business Units: Operations such as trust, accounting, and capital markets often run their own small data-processing operations. Usually, these functions were taken on because the departments were dissatisfied with the centralized data-processing department.
In some cases, departments can move or consolidate their computer operations into the centralized data center, but retain independent programming staffs. This way, the bank saves the cost of running redundant data centers while allowing departments to retain control over programming priorities.
* Cooperative Processing: A network of personal computers or advanced workstations can provide the same horsepower as a mainframe -- at a fraction of the cost. By migrating some functions from the mainframe to local workstations, banks can downsize the mainframe computers, along with the facilities and staff that support them.
* Case Technology: Consider computer-assisted system engineering a long-term investment, rather than a short-term cost cutter.
Case can be used to automate portions of the design, development, and maintenance of systems.
But some drawbacks: Programmers working with Case may require months of training -- and years of experience -- before they become significantly more productive.
* Image technology: Like Case technology, imaging may also require a substantial, upfront investment that takes years to recoup.
Item processing appears to be the most attractive image application. But the high cost of archiving digital images of checks makes it uneconomical to truncate the paper at point of entry.
Check-processing systems may carry a price tag from a few million dollars to $100 million. Another difficulty: The original checks must still be handled physically for statement rendering and archiving -- negating much of the potential savings.
A quicker return on investment may be a possibility via file-folder, or document, image processing. File-folder systems convert documents to electronic form and store them away.
This technology, which links workstations to a mainframe or minicomputer, tends to be more affordable than larger check-processing systems. File-folder systems, with approximately three workstations, start at around $250,000.
PHOTO : BRIAN WHITEHEAD noted that telecommunications, at some banks, are a tangle of inefficiency.