Printer Friendly

Five common computer management mistakes.

Personal computers have been with us for more than 10 years. Unfortunately many, or should I say most, organizations live in constant frustration with systems that either don't work or don't live up to their expectations. If your organization is in this bind, maybe it's because of some of these five common mistakes.

1. Making generalists out of specialists. In the mainframe world, functions are more specialized than with personal computers. For the most part, hardware experts work only on hardware. Software experts not only stick to software, but it is not unusual to find specialists who work only on limited applications. There are even programmers who make careers out of specializing in a single product from a single vendor.

In the PC world, the machines are less complex than their mainframe counterparts. The software on PCs can be just as complex as mainframe software. Yet it's rare to find the depth of specialization in PCs that is found in mainframes. People expect anyone who knows anything about computers to know everything. We ask programmers questions about word processing systems. Think about it: Why should a programmer know more about a word processing system than your secretary? But somehow we expect that this is the case. What's worse, I've seen things work the other way, too. I've seen people go to word processing experts for advice on systems. We confuse keyboard skills with computer literacy, computer literacy with expertise, and then wonder why the systems we build on bad advice don't work out. If you need computer advice, get it from the right source. (If you have a question about Lotus, ask a Lotus expert; if you have a question about word processing, ask a word processing expert.)

2. Failing to properly manage computers and projects. The single biggest mistake we make in managing systems implementation is premature implementation. I'm not saying we can't set aggressive deadlines in managing systems implementation; nor am I saying we can't meet the deadlines we've set. What I am saying is that systems projects are among the most complex activities organizations attempt. There is a structure and method to systems development and implementation that doesn't lend itself to cutting corners.

However, we manage systems projects like we manage people. Managers are aware of Pareto's Law, better known as the "80-20" rule. Applied to project management it's something like "80 percent of the work gets done in the last 20 percent of the time before the deadline."

This rule may work in other disciplines - although I doubt it - but in systems implementation it is a lightweight approach that leads to disaster. Imagine this worstcase scenario: A systems development company rushes its application out the door, and your organization purchases that application and implements it too quickly. You're going to just love the result, aren't you?

The best way to manage a computer project is to set realistic deadlines and resolutely attempt to meet them. But when it comes down to a choice between meeting a deadline and finishing software correctly, favoring the deadline leads to disaster. When it comes to software development, smart managers see deadlines as target dates, not drop-dead deadlines.

3. Maintaining unrealistic expectations of the software. Many users expect that the software they implement should cover 100 percent of their operational requirements. For example, we feel that a billing system should handle every exception encountered, no matter how insignificant.

I've sat through countless design meetings addressing questions like, "How does the system handle a member who sends in a check for dues and adds a payment for council membership plus a PAC donation but hasn't included enough money to cover the membership payment and is in arrears?' Illogical answer: Spend thousands of dollars and precious time programming software routines to handle this and every other nitpicking exception, even when you know you won't cover every possibility anyway. Logical answer: Pick up the telephone and call the member.

In other words, lower your expectations for software. If you purchase a system expecting it to do everything at the press of a button, you are bound to be disappointed. The happiest users are those who are self-sufficient people who want to get the most out of their systems and who are willing to work hard to get it. On the other end of the spectrum are users who bought a more powerful system than they needed and don't have the staff resources to properly use it. Most unhappy are users whose motto is: Our organization is unique. These people insist on customized software and expect systems to do everything just as their organization does it.

For best results, apply Pareto's Law with software functionality. Expect your software to cover 80 percent, or even 90-95 percent of your operational requirements, and handle the rest manually.

4. Cutting purchasing corners. Organizations on shoestring budgets would do better to stay out of technology than to buy cheap, off-brand hardware and hire low-priced, low-knowledge consultants to put together systems that are doomed to failure from the start. It's irresponsible to shop for hardware made in someone's basement when the top-of-the-line hardware is very reasonably priced. Ask any-one who sells the stuff: There is very little margin on hardware. If you find something priced significantly below the street price of a well-known brand, you can bet someone has cut corners somewhere.

5. Failing to standardize, sometimes referred to as inconsistency. The number one sign of an inefficient, poorly managed data processing system is a mix of different brands of hardware and several software packages designed to perform the same function (e.g., having several different spreadsheet or word processing systems in the same organization). Integration of information should be one of the goals of organizational data processing. PCs are great tools for an individual. But connect them together in an organization, and there are two things to remember: Personal computers are not personal and they are not computers. In an organizational sense, PCs are intelligent workstations on networks that serve the entire association. When your organization begins to talk about integration, you cannot expect to achieve your goals without standardization. The two are inseparable.

Standardization applies even more so to the data base system your organization selects. If one part of your association uses Dbase, another uses Paradox, another uses R:base, and so forth, you may appease individuals by allowing them to use their favorite system, but it's short-term thinking. In many respects, the product you standardize on is less important than the fact that you standardize at all.

There are a lot of other items that could make a list of the top mistakes: lack of adequate staff training; selecting hardware before selecting software; buying rather than leasing (or vice versa, depending on your financial religion); using pirated software; and ignoring security. The point is, if you are frustrated with data processing in your organization, these are some major areas you can look at to correct the situation. Steven L. Harrison is vice President of information systems at Electronic Really Associates, Overland Park, Kansas.
COPYRIGHT 1992 American Society of Association Executives
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Technology at Work
Author:Harrison, Steven L.
Publication:Association Management
Date:Jul 1, 1992
Words:1177
Previous Article:How to get fired without getting burned.
Next Article:Nonprofit publishing now.
Topics:


Related Articles
Human error responsible for drug injuries in hospitals, study says.
Do It Right the First Time: A Short Guide to Learning from Your Most Memorable Errors, Mistakes and Blunders.
NIST publishes computer security guidelines. (General Development).
I can't see you: getting the most from your off-site staff. (Inbox).
Outsourcing smarter: five outsourcing mistakes that could have a deadly impact on your company's bottom line.
Mining student data captured from a web-based tutoring tool: initial exploration and results.
Pain reliever: six common and costly management mistakes--and advice to help you avoid each one.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters