Zombie User Stories … User Stories that return from the dead!

Some User Stories that never seem to die. No matter how hard you work on them they get carried over to the next Sprint … Sprint after Sprint after Sprint. Welcome to the nightmare of Zombie User Stories(*)!

Zombie (User) Stories are a bad thing for a number of reasons; first, they sap the teams morale and reserve, and second, they quickly erode any trust that’s been established with the Product Owner. Eliminating Zombie (User) Stories is quick and easy following this simple three-step process:

  1. Write Acceptance Test for all User Stories. Without Acceptance Tests, how does a team know how much work to do? Quite simply, it’s impossible to know. And if you don’t know how large (or small) a Story is, then how would a team know if they can finish it within a Sprint? I’d recommend all Stories for the next two or three Sprints have Acceptance Tests defined before the Planning Meeting (here’s one way of achieving this).
  2. If a Story cannot be completed during a sprint, then break it down. If you’ve broken it down and it’s still took big to fit into a Sprint, then keep trying! And,
  3. At the end of Sprint the Product Owner should simply reject all stories that are not done. Product Owners can help by rejecting any User Story that might be only partially completed. This will help the team understand what “Done” means, and by consistently doing this the team’s behavior will change.

Now that you what to do, get out there and put an end to those Zombies User Stories!

(*) I admit that I often overdo the Zombie theme, but that’s because of my “Zombie survival leaflet” exercise during my courses. The most fun you’ll have with braaaaiiiinns!

, , , ,

One Response to Zombie User Stories … User Stories that return from the dead!

  1. Felix May 18, 2011 at 9:29 am #

    Hi,
    I agree. Usually Zombi User Stories get into the Sprint when they are not yet READY. As they are not ready to be DONE, thus keep on lurking in the Sprint Backlogs …
    Felix

%d bloggers like this: