Time Boxing
Evolutionary Project Management (Evo)
uses Time Boxing rather than Feature Boxing:
- Feature Boxing: Waiting until the work is done
- Time Boxing: Defining just enough time to do the work as needed. There is no more time
TimeBoxing provides incentives to constantly apply ways to save time in order to stay within the TimeBox.
TimeBoxing is much more efficient than FeatureBoxing (waiting until we're ready),
because with FeatureBoxing we lack a deadline, causing
Parkinson's Law
and the
Student Syndrome to kick in badly.
Note that this concept of saving time is similar to "eliminating waste" in
Lean thinking,
and already indicated by Henry Ford in his book "
My Life and Work", back in 1922.
In the Evo TaskCycle we learn to optimize our use of time ever better, using TimeBoxing.
In a project people didn't use TimeBoxes and every time I suggested using them,
the PM said: "There's Niels again with his TimeBoxes, they're not efficient!
If at the end of a TimeBox someone is almost done with his job, should he drop his work? That's inefficient!"
"Yes, he should drop his work!" "But that's inefficient!" "That's true." "Then, that's not a good thing to do!"
"Yes, it is a good thing to do. Just because it's inefficient it creates pain. And with pain we can learn faster..."
I found that if we are serious about time, normal people need about three weeks to change
into realistic estimators, being able to estimate a TimeBox in which they can do the work.
Besides, I don't even mind whether they keep their TimeBox or not. The whole estimation exercise
is not to come up with good estimates. After all, estimates don't pay salaries. Only Results do.
It's just to learn how to promise what you can do and
then to live up to your promises. And, as said, it takes only about three weeks for people to learn.
A TimeBox is the maximum time available for a Task.
When the time is up, the Task should be completely done: there is no more time!
Because people tend to do more than necessary (especially if the requirements of the Task are unclear),
there is usually ample opportunity to conclude the Task within the TimeBox:
- Check halfway whether you’re going to succeed on time
- If not: what can you do less, without doing too little
- Define the requirements of the Task well
- If the TimeBox is unrealistic: take the consequences (pdcAct) immediately
(if a Task suddenly proves to need much more time, is it still worth the investment?)
If you really cannot succeed within the TimeBox:
- Check what you did
- Check what you didn’t do
- Check what still has to be done
- Define new Tasks with estimates (TimeBoxes !)
- Stop the current Task to allow for finishing the other committed Tasks
(don’t let other Tasks randomly be left undone!)
- If you still have time left after having finished all Tasks for the week, you still
can try to finish the uncompleted Task
Note: Don't believe anything I say! You can use my advice, but
it's your success that's on the line! So, ultimately you should find out yourself what works best for you.
Plan-Do-Check-Act is the magic technique to find out quickly.
