Keeping Performance up to Speed


Last week, while I was waiting for a phone call at my home office, I ran a utility program on my relatively new computer for the first time. I purchased the system last spring, and while I decked it out with all of the appropriate antivirus and automatic update features, I hadn’t yet run a maintenance check. After all, I’ve only had the computer for a few months, well, o.k., it’s been seven, but what could possibly happen in that short amount of time?

Apparently, plenty. The software found one what it called “major” problem and dozens of minor ones. My system was “moderately fragmented,” which meant that the computer had to search through the disk to find different parts of a single file, an inefficient process. No problem—that’s what maintenance programs are for. It fixed the errors, defragmented my hard disk, and I was back in business.

What I didn’t anticipate was the radical improvement in the computer’s performance after I had done this housekeeping. It blazed! Programs launched in the wink of an eye, graphic-heavy web sites loaded in an instant. As I witnessed these feats, I was reminded of my amazement at how speedy the processor was when I first plugged the computer in.

The question that puzzled me was, why didn’t I notice the computer’s performance had degraded so much? After thinking it through, I concluded that, little by little, I had shifted my expectations. The decline had been gradual, and performance was still within acceptable limits, so I easily adapted to the slower access and load times. However, if I had continued to put off the maintenance process, the computer would have eventually crashed, which certainly would have gotten my attention and caused untold difficulties.

In this case, the consequences were reversible—I was quickly able to fix the system so that it ran as efficiently as ever. But when this dynamic occurs in other situations, it can be more difficult to diagnose and the results can be more damaging.

Lowering Performance Goals

In systems thinking terms, I had experienced an example of the “Drifting Goals” systems archetype. Systems archetypes are common patterns of behavior that occur in all kinds of settings. “Drifting Goals” involves lowering our performance goals rather than

Check-ups or maintenance programs use objective measures of a system’s performance to periodically diagnose problems that might not be apparent to someone on the inside.

taking corrective actions. Sometimes we do so because these actions are undesirable, as in the case of cutting expenses in order to reach profit goals. Sometimes we’re focused on other factors that seem more important; for example, we may be so caught up with efforts to boost sales that we fail to notice that quality has slipped. And sometimes, as I experienced with my computer, because our senses aren’t attuned to gradual changes over time, we just don’t notice that performance has degraded.

The parable of the “boiled frog” is often cited as an example of the “Drifting Goals” dynamic. According to the story, if you toss a frog into a pot of boiling water, it will immediately try to jump out. On the other hand, if you put it in cold water and then gradually raise the temperature, the frog will happily swim around until it—there’s no delicate way to put this—cooks. The frog’s survival instincts are geared toward detecting sudden changes, not incremental ones. Although this fable has been questioned by scientists, it vividly illustrates what I experienced with my computer —I likely would have noticed an abrupt decline in functioning but was unable to detect a slowdown over several months. Just as the frog adapts to the water temperature, I unconsciously lowered my expectations of the computer’s performance.

Adjusting our expectations isn’t always bad, but if we’re going to change our goals, we should do so consciously. The key is to know what our objectives are and to track performance vis-à-vis these benchmarks. To that end, most manufacturing companies have mechanisms in place for monitoring adherence to quality standards. Organizations also tend to stay on top of financial and sales goals through routine reporting and analysis.

When it’s not practical to measure performance on a continual basis, as with my computer, then a regular check-up may be in order (see “Maintaining Performance Goals” on p. 8). Check-ups or maintenance programs use objective measures of a system’s performance to periodically diagnose problems that might not be apparent to someone on the inside. For instance, in a physical examination, a doctor checks blood pressure, weight, cholesterol, and other levels to ensure that they remain within healthy limits. Unless a person has a health problem that requires continual monitoring, such as diabetes or high blood pressure, checking these functions daily or weekly would be onerous—for most people, once a year is often enough. But if we put off seeing our physician for too long, changes that we’ve gradually adapted to—low-grade fatigue or a persistent cough— may compound to become health crises.



In certain systems, such as my computer, actual performance begins to fall short of desired performance over time (B1). We may not notice the shift, because our senses aren’t attuned to gradual changes, so we unconsciously lower our expectations of the systems’ performance. Rather than changing our standards, a more productive approach is to consciously be aware of this dynamic and to institute a regular check-up or maintenance process (B2). By doing so, we bring actual performance back up to speed and keep our goals on track.

Organizational Check-ups

How might we incorporate the idea of maintenance checks in an organizational setting? The goal is to cast light on changes that we may not perceive because they are so gradual while not becoming bogged down by burdensome reporting or monitoring tasks. This is especially true for areas that aren’t easy to measure, such as employee satisfaction, adherence to the corporate mission statement, or teamwork. In these cases, a maintenance process may be as simple as meeting with a partner every week to get an objective opinion of your progress on achieving a developmental goal or as complex as conducting quarterly employee surveys to evaluate morale.

Here are some ideas for making sure that performance stays steady over time:

  • Identify variables that are important to organizational performance, especially those that aren’t usually on the radar scope, such as employee morale or use of productive conversation tools.
  • Establish performance standards for these variables. Keep the standards visible.
  • Track performance versus the standards.
  • If it’s not possible or practical to track performance analytically, find a way to periodically collect input from an objective source—a learning partner, an outside coach or facilitator, a semi-annual employee survey. Experiment to find the right interval between “check-ups”—too often and you might find them more trouble than they are worth, too infrequent and problems might be on the verge of spinning out of control before you catch them.
  • If you are tempted to shift a goal, be deliberate! Look into the causes and consequences of doing so before taking action.
  • Learn from experience. If you’ve noticed unacceptable changes in a variable, design a maintenance program to keep it on track in the future.

Resources on the Systems Archetypes

Systems Archetypes at a Glance by Daniel H. Kim

A Pocket Guide to Using the Archetypes by Daniel H. Kim and Colleen P. Lannon

Systems Archetype Basics: From Story to Structure by Daniel H. Kim and Virginia Anderson

Applying Systems Archetypes by Daniel H. Kim and Colleen P. Lannon

Systems Archetypes I: Diagnosing Systemic Issues and Designing High Leverage Interventions by Daniel H. Kim

Systems Archetypes II: Using Systems Archetypes to Take Effective Action by Daniel H. Kim

Systems Archetypes III: Understanding Patterns of Behavior and Delay by Daniel H. Kim

These and other resources are available through


Please send your comments about any of the articles in THE SYSTEMS THINKER to We will publish selected letters in a future issue. Your input is valuable!

If you follow these steps, you’re likely to keep your organization performing at high levels and avoid crashing the system or boiling the frog —things none of us want to do!

Janice Molloy is managing editor of The Systems Thinker and content director at Pegasus Communications, Inc.

Sign up or sign in to bookmark this article.