Fear of Measurement followup


A few weeks ago, in this post, I asked the following question:

Consider the following metric, then tear it apart and make it better. Tell me what could go wrong with it, how to make it more accurate or actionable or any other way it could be improved to be more beneficial.

% of tasks planned versus completed for last month

Despite the weaknesses I’m about to point out, it’s similar to goals I’ve seen in organizations recently, and probably similar to what most of you see on your reams.

Discussion Points:

Could encourage team members to add unnecessary “buffer” to the schedules

What happens to tasks that weren’t completed last month? Do they get added to the work for the following month, or are they never counted again?

How do you handle the case of tasks from next month being completed this month – i.e. what if the team “works ahead”?

How is the data collected – i.e. is project management consistent?

Does the metric help us understand the overall goal (Improve the accuracy of estimates for all projects by 20% in the next year)?

_________________________

You don’t need answers to all of these questions in order for the metric to be useful. However, you do need to think about things like this any time you are going to start measuring something (and yes Geoff, you need previous data too).

I’m just saying that as with most things in life, common sense prevails over blind faith. ymmv.


Comments (2)
  1. James says:

    I think the "buffer" question is best covered by having solid test entrance and exit criteria or well defined start and complete milestones.

    The joy of where I work is that new requests for unplanned work, SW release slips, and underfunded or unfunded "accepted" tasks always make it a challenge to properly plan out work for the long term.

    But, I guess that’s why I’m a manager….

Comments are closed.

Skip to main content