Benchmarking for the Rest of Us
There is no shortage of data available to determine whether your software team is sufficiently productive. Whether it is the often quoted Chaos Report from the Standish Group, the quarterly updates from the Software Engineering Institute or the hidden project data behind parametric estimation models such as COCOMO II, it is seductive to hold your own performance against these standards for comparison.
We quickly find, though, that these comparisons bear little relevance for most organizations, especially those that are ‘up-and-coming’. It is the growing companies who are most in need of benchmarks to gauge their performance and progress. This discussion identifies the challenges with most published information, and enumerates the approaches that we can all use to generate meaningful benchmarking information.