Metrics are a double-edged sword. On one hand, you can't improve what you don't measure. On the other hand, if your metrics program is implemented poorly, you can end up with all sorts of problems ranging from fooling yourself into thinking everything's ok to demotivating your team to information overload. In any product development environment, you really need to pay attention to two classes of metrics:
- product metrics - tell you about the quality of your product
- team metrics - tell you about the productivity and effectiveness of your testing team
If you're putting together a metrics program for your product, here's a short list of some articles I found useful while working on ours:
- OntoInfo.com presented some great thoughts last Fall on software metrics and how to make them actually work for your team - boiled down to 7 simple points.
- "Measurement Issues and Software Testing", by Cem Kaner
And if you're designing a metrics program to help increase the productivity of your team, I recommend:
- "Measuring the Effectiveness of Software Testers", by Cem Kaner
- "The Seven Habits of Highly Effective Testing Organizations", by Lee Copeland
Of course, these are just a start, but they sum up the key points quite nicely and can help you avoid common pitfalls.