Tuesday, October 4, 2011

Ode to ETC – The Task Overrun Early Warning System

History is a vast early warning system.

                                                - Norman Cousins (15Apr1978)

Sometimes you have to use the tool you have available, whether it is any good or not.  So it is with getting task completion status as percent completes.  Much better though is using the best tool for the job.  This blog exists to share project management best practices, so today I extol the virtues of ETC.
As PMs, our job includes delivering a project according to plan, knowing when the project is off plan (and acting to get it back), and communicating status to stakeholders.  Other people – our project team mates – do the important work and we rely on them for the information to do our job.  To be successful, we need a tool that helps our team mates communicate that information and for us to receive it accurately.

The well-defined task is the starting point.  Knowing whether the task has not started, has started (is in progress), or is completed is beneficial.  It helps us with history, with past events, but doesn’t communicate enough about the future.  Looking back is easier and more comforting, but it’s not the most beneficial nor the most important activity.  The most important objective with task status is to look forward, to be able to determine that the task will progress to plan.  Only by looking forward effectively are we appropriately serving our stakeholders.
The best tool for tracking progress is Estimate to Complete (ETC).  ETC works best when tasks are estimated, scheduled and reported by effort (as they should be), but even with duration scheduling is still superior to other methods of communicating what remains to complete the task.  When you ask the task owner “How much more work is (or how many more hours are) required to complete this task,” it forces them to re-estimate the remaining work, but with the advantage of everything they’ve learned by progressing to the current state.  The values get progressively more reliable.  Further, as the task gets closer to completion and the ETC value gets smaller, the task owner has an incentive to complete the task (just to get it off their plate), as opposed to the negative incentive system with percent complete reporting (see my next post).

No predictive system works effectively if it is rule based.  That is, if the task was originally estimated at 40 hours and the owner has completed 24, they can’t satisfy the ETC just by reporting 16 hours to go;  that defeats the predictive value of the technique.  They have to actually re-estimate the remaining work, which could be more or less than 16.  Tip:  be very suspicious if a task owner just keeps reducing ETC by the number of hours worked on the task.
ETC, when used consistently and properly, is a PM’s best friend.  ETC is an early warning indicator into task delays, under-the-cover scope creep, or a task owner that may be over their head or not performing.  As an early warning indicator, ETC is much more reliable than other methods.

One thing I’ve found using ETC is the necessity to frequently reinforce that the task owner must re-estimate the task each week.  Do you have any examples of problems using ETC?

No comments:

Post a Comment