Re: Dev Teams Hobbled by Poor Metrics, Study Suggests
I tend to think of project metrics and quality metrics as separate spheres. We have lots of ways and tools for collecting quality metrics through the SDLC, measuring effectiveness, coverage, defect discovery rates, JiT quality, etc. Project metrics are assessing progress against plan, actual against historical or estimated rate/speed.
In reading the article, it seems the focus is more so on project than on quality metrics.
The article's first suggestion of collecting metrics iteratively does little to decrease the perceived 'cost' of metrics; merely decreasing the interval while increasing the frequency - and possibly improving accuracy somewhat (through timeliness).
The second suggestion - 'practice a "disciplined estimation of business value" in their project estimates' - is never explained, just asserted.
My own take would be to have some method of estimating and then tracking business value-add for each discrete development object and/or task. That can be inclusive of the QA/testing function.
Re: Dev Teams Hobbled by Poor Metrics, Study Sugge
Good take on it bkoenig. It is kinda a lame article. I guess the spheres overlap here and there. Some of the same techniques and process would be used in both. It's the specific data/metric that's different.