Getting things done + the expected outcome of code quality/testing (rather than that as an end in itself). Of the two metrics we've recently started looking at,

Getting things done + the expected outcome of code quality/testing (rather than that as an end in itself). Of the two metrics we've recently started looking at, #2 tries to capture this. If we can reduce it without resorting to some prescribed form of test coverage, it's perfectly fine. I think. 1. Deviation from original estimate as a percentage 2. Time spent on bug fixing as a percentage of time spent on new development Still new, so this could go either way...