Printer Friendly

What should a CFO know about dispersed computing?

What should a CFO know about dispersed computing? Information systems have become a competitive resource in many organizations. Now, financial executives responsible for IS have to worry about how to provide effective systems to their many and varied clients within the organization. Is dispersed computing the answer? As CFO of a medium-sized commercial bank, Steve Dell has the information system (IS) function reporting to him. As he went through his mail one Monday morning, he found an astounding memo from Lew Dollar, the vice president in charge of the cash management function. It went like this: Dear Steve: As you are aware, my function is coming under increasing pressure from our competitors, who are promising far higher returns on their customers' accounts than I can and who are providing customers with online access to their accounts on an as-needed basis. In this division, we have lost three of our largest accounts in the last six weeks; the clients point to our poor cash management capability as one of their chief reasons for switching.

You have had about a dozen people in your IS group working on our new cash management system for more than a year. My people have devoted a great deal of time to specifying our requirements for a competitive system and have been charged back more than $1 million on this project to date. Vic Clarke, our account manager from IS, told me last week that the project could be completed in "about a year," but this is the third time that the target completion date has been revised. I have no confidence that it will not be changed several more times before we have a functioning system. Given the competitive pressure I am experiencing, if I do not have some new capabilities in the cash management area within the next six months, we risk losing significant share in this important market. Bottom line, I will not be able to meet my business plan contribution to the bank's profit margins.

Therefore, I am initiating action to examine cash management software from outside suppliers, possibly as a stopgap measure.

At present, your people's careers do not depend on my business success. If we had IS people reporting directly to me as an integral part of my staff, I believe they could find ways of getting better systems faster. Therefore, I am requesting that you transfer your 12 IS people, plus a manager, to my organization, reporting to me, effective 30 days from this date. I will provide access to whatever technology and space they might need. Permanent assignment of IS people into my area will greatly expedite systems development since these staff members will develop an excellent understanding of my side of the business by being a part of it.

Steve, I am hoping we can arrive at an amicable agreement on this and that I will have your full support. In any event, I am prepared to carry it to whatever levels necessary to get the resources to regain our competitive position. In essence, Lew wants to control his IS future in order to control his own bottom line. He can't wait the usual three to five years for his new system; he may not be in the cash management business by then. He is betting that if the IS people report to him, they will identify with his success and do a more effective job in meeting his competitive needs. He may not care much about how cost-effective it is, as long as they do a good job on time.

The case illustrates the three major themes in managing IS today: * Capability--Can our IS effectively meet competitive threats? * Cost--Are there economies of scale in centralized IS, or has the overhead and chargeback gotten too high? Can we downsize systems to cheaper technology? * Control--Should we locate the systems and report on the data where the business needs it?

What capabilities are gained? Lew's memo on IS dispersion reflects just one of many methods for improving IS capability in a specific part of the business. And these pressures exist not just in banks, of course, but in all industries where information systems can be used to gain competitive advantage.

The theory behind dispersion runs something like this: in situations where information systems can provide competitive advantage, there has to be a free interchange between the businesspeople, who see the competitive opportunity, and the technologists, who know how to build systems to do something about it. The best way to do this is to locate the technologists in the business unit. The interchange has to be informal and day-to-day. Ideally, technologists must have their professional success linked to the same business objectives as the businesspeople do. In most of these situations, the location and ownership of the processing capability is irrelevant.

There are strong pressures for dispersion in about a third of the companies we looked at in a study sponsored by the Society for Information Management (SIM). In these high-dispersion companies, the systems development function and more than half of all IS professionals are located in and report at the business unit level or lower in the organization.

The driving force for IS dispersion in these companies is the increasing competitive use of information systems. In these cases, the capability of the IS becomes more important than the efficiency and the careful control of IS costs. This is because the primary focus has changed from cost savings to new revenue generation. The two objectives have equal profit impact. Cheaper technology has made dispersion a more practical alternative, and increasing dissatisfaction with IS performance and chargebacks has added fuel to the dispersion fires.

So in these situations the traditional beliefs about IS economics--that there are economies of scale and that shared resources are more cost-effective than dedicated ones--are coming under attack. Centralized IS is seen by business units as unresponsive, unable to understand the business, and too slow and too costly. Competitive urgency is driving out all of the traditional concerns about IS efficiency. And business unit managers, who see their fates tied to IS effectiveness in meeting competitive threats, want more control over their IS capability.

But what really happens to IS effectiveness when IS resources are dispersed? According to the SIM research, no one knows for sure. Dispersing companies in the study found the transition painful, but most would not go back to traditional ways of organizing IS. Like Lew Dollar, they were betting that realigning IS closer to the business would improve its effectiveness. At this time, a new research study, endorsed by the Committee on Information Management of Financial Executives Institute, is being launched to develop measures of IS effectiveness and determine its relationship to IS organizational alignment strategies.

Where does cost come into play? The drive for increased IS capability can run smack into senior management concerns about costs. In one large bank, the annual systems development agenda and budget is controlled by a committee consisting mainly of business-unit executives and chaired by the bank president. IS costs are a major component of non-interest expense for most banks; so they also are targets of opportunity when executives seek to improve the numbers in the annual report. Two years ago, the agenda was completed after much struggle and argument, after which the president, along with the bank chairman, requested a 20-percent cut in the development budget. With much grumbling, this was arranged. The following year, when the same request was made, business-unit executives called another meeting of the committee and announced that, if the cuts were again implemented, they could not commit to their projected bottom lines. The cuts were rescinded.

The key event here was that, between one year and the next, responsibility for constructing the case for new systems was transferred from the systems department to the business-unit managers. Traditionally, the case for systems was based heavily on the reduction of staff costs and other efficiency measures to be attained through automation of office processes. It turned out that, in presenting their cases, business-unit managers emphasized benefits from added revenues. This would be achieved through providing better and more differentiated products and services, rather than just cost savings. But to meet the new revenue targets, these managers needed new and effective information systems, which meant they had to gain control over the systems development agenda and budget.

Who should control IS within a company? According to our research, IS dispersion often begins more like a social movement than a formal organizational decision. It may take place over several years, one event at a time, with the final result being increased user control of IS activities. Here are some key indicators of the pressure to disperse: * Business units are gaining influence over the systems development agenda and/or budget. * "Bootleg" development shops are springing up in business units. * There is increasing clamor from business units about IS costs, poor delivery, and so forth. * Business units are stepping up their purchases of outside software for either business-critical applications or to meet competition. If any of these indicators exists in your company, the movement for user control may already be started. The result is usually IS reorganization, and the dispersing of IS resources outward and downward in the organization. If senior IS management resists the pressure, a common response is to replace the top IS executive with a non-IS businessperson who implements the change.

If such user control of IS is a solution, what problems does it address? First, it can provide business-knowledgable service to the different IS clients. It also can manage the tradeoff between efficiency and effectiveness. Because the business-unit manager has responsibility for both the cost and the revenue, the tradeoff decision between cost savings (efficiency) and revenue generation (effectiveness) is made at the appropriate level.

Further, user control ties IS staff identity, goals, and rewards to the business-unit success. It breaks down the cultural walls often thrown up between IS and the business unit. And a working partnership between IS and the businesspeople replaces the traditional service-bureau mentality.

Finally, user control moves IS problems away from senior management attention.

There are some risks to dispersion, of course. Primarily, Lew Dollar risks losing some of those new IS professionals, if they feel isolated in the business unit or if their chances for learning, growth, or promotion are slimmer in the new organization. There is a learning curve for the business manager who begins managing IS professionals. For example, IS professionals depend upon high levels of training in order to stay technically current. Lew must also invest significant time and money in expanding their business knowledge and communication skills, if he expects them to meet his expectations. He also has to open new career paths that motivate these people to expand their skills beyond technical expertise.

The best chance for IS dispersion to succeed occurs in a decentralized organization that hires generalists and then lets them rotate through a number of business and functional units before they settle on a career line--and this could include IS. But it is a long-term cultural change that usually requires five to six years to get in place.

Dispersion can be implemented with good planning and can avoid drastic cultural change. Which IS functions should be dispersed? Good candidates are business systems analysis, systems development, and end-user computing. Poor choices are database and network architecture (including LAN standards), systems standards, and systems security and audit.

Another important question is to whom the IS resources should be dispersed. Among the possibilities are business units, profit centers, or departments needing better systems to meet competitive challenges. Another choice is subunits whose information needs or other work does not depend on anyone else.

How fast should the information systems be dispersed? Time is needed to plan the move and to select and prepare both the movers and the receiving organizations. Human resources support must be beefed up during the transition. And units should proffer common pay scales and career ladders for the IS people.

Alternatives for IS realignment Of the various alignment strategies, IS dispersion and decentralization are the most popular at the moment. But companies are also experimenting with setting up IS as a profit center within the company, or spinning off the IS organization as a subsidiary corporation that is responsible for its own bottom lines and is free to accept work from sources other than the parent.

Decentralization of IS usually means breaking up a large centralized IS group and transferring some functions to business units, divisions, or departments--usually as intact groups with organizational walls between them. Be aware that this often prevents the blending of business and IS cultures and the sharing of essential knowledge. In effect, you can end up converting one big problem into a number of smaller ones, with the decentralized IS people more isolated than ever.

Other dangers of dispersion include losing control over data and technology architectures.

In banks and other financial services organizations that depend on access to core systems, loss of architectural control can spell disaster. Indeed, four years after one midwestern bank dispersed computing completely and gave up architectural control, it found it was unable to build "customer relationship" files cutting across lines of business. When the bank saw other banks touting this as a customer advantage, it recentralized architectural control but only after losing several years to the competition.

Under profit-center management, the user is free to negotiate the terms of development projects with IS, but IS doesn't usually take outside work. If IS "makes money" on the contract, it keeps it; if it loses money, it eats the difference. Such a system can increase reward congruence between IS and the business if both sides are good negotiators, but probably does not aid the cultural blend.

With the IS spinoff, both sides are free to do systems business with whomever they please. But what sometimes happens as senior management spins off IS is that the user community reacts by building its own in-house IS staff (to avoid paying "profits" to the spinoff company). And the spinoff company suddenly finds it lacks the marketing skills to compete with the systems houses of the world. Some spinoffs completely sever their relationships with former corporate parents. The EDS-General Motors merger is an extreme example of an IS spinoff.

All of these organizational strategies can hide IS issues--and the total IS budget--from senior management attention. Indeed, under some organizational alternatives it is difficult, if not impossible, to identify the total corporate IS expense, since it is being managed at lower levels and it is difficult to identify IS people or technology from ordinary business or office expenses.

Still, many financial executives, like Lew Dollar, are betting that realignment really can improve IS effectiveness in meeting business challenges--as long as the systems are thought of as essential to business performance rather than as corporate overhead.
COPYRIGHT 1989 Financial Executives International
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:dispersed computing in information systems management; chief financial officer
Author:Lodahl, Thomas M.
Publication:Financial Executive
Date:Jul 1, 1989
Previous Article:The new dilemma of cash versus earnings.
Next Article:DEC's decentralized financial system puts strategy above controls.

Related Articles
What's your IT score?
What's to Become of IT?
IT De-volution.
Front Lines.
When police might use force in crowd control.
Mixing/dispersing systems.
Technical committee profile: CFIT; Committee on Finance and Information Technology.
8 arrested after fight outside nightclub.
CFDCs bolster job growth.
Tough threads.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters