3

I have a team of 7 developers and we are doing a software maintenance project for our client.

I want to give objective feedback to my subordinates and will like to include productivity as one of the parameters in the feedback.

What is the best way to calculate individuals productivity? I was thinking of using FP (function points) per week as a measure.

Is there any other industry standard which I can utilize?

4

I don't know if it is a standard, but probably the best way to assess productivity is to track the number of deliverables completed and accepted by the client. The advantages to this approach are:

  • Truly objective and simple to measure. Something is either done and accepted by client or it is not. And you avoid the risk of something sitting around being 90% complete for a prolonged period of time as the developer tries to fix that one last thing.
  • Incorporates quality of work done. By having the client accept the deliverable before it is considered complete you avoid the risk of incenting your developers to churn out code without considering how well it works.
  • From the client point of view it is the only measure that matters. By paying attention to this you integrate productivity measures with customer satisfaction, which is presumably a key business objective.

The primary disadvantage with this approach is that someone with a very complex deliverable could appear to be less productive than someone with several very simple deliverables. You avoid this if you look at team productivity as opposed to individual productivity.

  • We currently are experimenting with two approaches. First is to come up with FP estimate (Size) for each change. Problem with this approach is we dont have enough expertise for FP estimation within the team. Second is to define SMC guidelines for every type of change we do. And assign weight to the SMC guidelines. But the SMC guidelines have to be updated frequently (as per new changes being introduced). So for every change delivered to client is categorized in SMC and time spent on each change is measured. Thus we arrive at productivity number. – ViSu Apr 30 '14 at 4:10
  • Our second approach is in line with Doug's suggestions, but I was not sure if its the most scientific way hence asking the question here – ViSu Apr 30 '14 at 4:14
0

I will second Doug's answer of using deliverables as the metric, but add one additional caution point.

If you're rating someone on productivity, you (and the team member) need to be clear on what the baseline is - productive based against 'what' expectation? For this you and the team member would need some estimation of how long the deliverables/tasks should take, so you can then measure against these expectations.

This also allows for the differing complexity that Doug mentioned.

So my suggestion is let them know that this will be a rating metric, and then work with them to establish the expectations, and only after that begin the measurement.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.