---

Metrics are merely measurements; i.e., methods by which to measure something. Within the software industry, such metrics have been around for many years and used in ways that enhance efficiency. Despite this, there are companies that fear metrics. This is probably because of the frequent and egregious misuse of metrics by these companies in the past. The misuse of metrics can destroy good processes and perpetuate poor ones, as well as improperly reward or punish people.

If your company wishes to put stress on your project team and instill fear into them, here are five ways to misuse metrics:

Misuse #1: Select Metrics Arbitrarily
Before selecting metrics, it's considered wise to understand the process that you are measuring as much as is possible. In that way, you can select metrics that best measure areas meaningful to your project team. Making metrics up just to impress upper management is a good way to create a lot of meaningless numbers that are time-consuming to track and add no value to the project because they do not indicate a logical direction in which to go to improve.

Misuse #2: Don't Ever Change the Metrics
It's not possible to select the perfect metrics on the first try. In fact, it's just not possible to have a perfect set of metrics. However, the ones a project team comes up with first are not going to entirely measure what they were intended to measure. Surprising results tend to come out of some of the metrics. Sometimes, the surprise is that your situation is different from what you thought it was. Other times, it means the metric is bad. If bad information is desirable, just keep measuring all these same items, over and over.

Misuse #3: Measure Everything
One might consider that everything worth doing is worth measuring. However, it is not the case with everything we do that it is particularly useful or interesting to measure. So, if the idea is to bog-down the project team to the point that they spend more time measuring things than actually doing things to be measured, then please go ahead and track everything. But if the purpose is to identify what, where, and how to improve, then find ways of measuring only meaningful things.

Misuse #4: Perform All Measurements on the Same Schedule
Different measurements are useful in different increments. Truly, we cannot find the ideal increment for every individual metric. If we sent out one metric per report, we'd overwhelm people with reports. On the other hand, we can also trivialize our metrics by forcing them to operate on the same schedule. Some metrics are needed daily, such as what products made in the last 24 hours are out-of-spec. If you sent out that report only every week, the out-of-spec material is already in the hands of your customers.

On the other hand, if you want to overwhelm and bore people every week, send out huge turnaround reports that management usually wants to see only monthly. You could also try to send everything out at least daily. After all, they're automated, no printing cost is incurred, and more is always better, right? In this way, the metrics that are needed daily get lost in an overwhelming amount of material so nothing gets read. And, by the way, that out-of-spec material gets shipped along with the good stuff.

Misuse #5: Let the Metrics Drive the Process
Rather than creating metrics that measure process performance, let the metrics drive the process instead. If your process had been fairly decent to begin with, rearranging it to fit some arbitrary metrics is a good way to make it inefficient. If you select only one terrible way to use metrics to kill your process, this is probably it.

Finally
Once again, there's just no better way to demoralize your project team than by holding them to inappropriate metrics. When you implement any of the methods described above, the motivation and productivity of your entire project team can be destroyed.

---

---