I just read this recent article from McKinsey and Company that extols the virtues of software productivity metrics, compares several alternative methods, and offers their recommendation.I find it amazing that in the several years I have been publishing this blog, I have never discussed software productivity metrics – but after a review today, that is exactly what I found. As an avid advocate and proponent of metrics, the topic has certainly been on the list for discussion.
Had I written the article before today, I would’ve advocated Function Points. Function Points have several advantages, including being the most consistent, accurate and reliable of the various methods and the one most appropriate for evaluating and measuring vendor software. In particular, if you are comparing building a package versus buying a commercial (COTS) package for the same capability, then FPs offer an objective measure for comparing both features and cost to determine the appropriate path. Likewise, it is similarly appropriate when comparing products from multiple vendors.As the McKinsey article explains, though, FPs are expensive and challenging to introduce. For organizations that rely primarily on software development (rather than purchasing vendor offerings), the McKinsey authors introduce a new (to me) method called Use Case Points, compare this to Lines of Code, Story Points, and FPs, and describe how to introduce UCPs into the organization. Further, the article notes that UCPs are comparable across applications and teams. And UCPs can be used with both agile and waterfall development methodologies.
I encourage you to take a look at this article for your application development environment. Metrics add a valuable tool to your toolbox. And metrics are an absolute, foundational necessity for continuous improvement.Have you worked in an organization that uses software productivity metrics? What benefits did you find? What problems did you encounter?
© 2013 Chuck Morton. All Rights Reserved.