Dave Nicolette posted an interesting blog post about using function points (FP) as a means to compare productivity between IT projects using different processes. I understand the concept – find a metric that speaks to the audience we are trying to communicate with that allows an apples-to-apples comparison. However, why do I find this still an unsatisfactory answer to the productivity issue?
I tend to think this metric would only be useful for academic types or a research paper. Sure, it is create more evidence that “Agile is great and waterfall is bad”, and how long do we have to wait for this evidence? One business cycle? Since I am on a roll smashing your good idea, I also think the barrier for entry is too high for people wishing to make the case Agile would be good for their organization. Not only do they need to understand Agile practices & techniques, they now need to become FP counters. Now we need to start counting FP on all our IT projects? Is this something our industry has the skill to do? The evidence I have seen shows we do not.
Yes, FP are “good” because ISO has made them the standard, there is clear way to count them (for the people who know) and academics (and attornies) love them. There are real reasons why FP are not used in US business today. I work with the ideal organization (large Fortune 500 company) for FP analysis, they are always hungry for metrics and they do not use FP. I wonder why that is the case? It is not due to a lack of education. There are well-placed, knowledgable individuals aware of this metric, so there must be something else preventing FP. We need to understand why people who know about this metric, can afford the consultants to do the analysis and still do not use the metric. When we overcome those issues, then let’s talk about pushing this metric on Agile teams. In the meantime, let’s stick with “basics” like throughput, takt time or use EVM (as much as I hate it – it is comprehensible to business folks).