Dealing with metrics is one topic that is going to come up some time in your career. At some point, your boss or higher-up will request metrics from you or your team.
Before we begin looking at the types of metrics, let’s set the playing field. There are two things for you to remember. First, metrics without interpretation are useless; and second, the person seeking metrics must come clean as to what they are looking for. These two pieces of information will determine whether or not your metrics gathering will succeed or fail.
Development metrics may be the most controversial of the three. We all know about the uselessness of lines-of-code (LOC) metrics. Some developers figured out that if they just hit the “enter button” more, they could dramatically increase perceived output. Java.dzone.com gives us a way to look at development metrics from an agile perspective, and a blogger on Build Real Software offers an opinion on what development metrics we should be gathering.
Testing, sometimes mislabeled as quality assurance as I found out, also has a great need for metrics. I found a great website for those in the beginning stage of gathering the correct metrics for testers. Softwaretestpro.com provides a plethora of links on testing metrics. Askcts.com also has a good whitepaper on testing metrics that you may find useful.
Let’s take a look at configuration management metrics. Variety News provides twenty-five different configuration and change management metrics. Frank Watts, a frequent contributor to CM groups on LinkedIn, has written a book on CM metrics. You can view some of the chapters on Google Books. For those looking for an ITIL slant to configuration management, simplicable.com has metrics for you.
There is a lot of information to digest when it comes to metrics, and we didn’t even look at other areas in IT where metrics can be measured, like SaaS or IT service management.
The main thing to remember when looking for the right metrics is to find out the who, what, when, where, and why of the asker and the audience.
What has been your experience with software metrics? Has it been pleasant or miserable?
