When evaluating a sprint, the sum of all user stories that are DONE, give us our sprint velocity. Now, what if one user story didn’t make it to done, but is finished for 90%? Do we add 90% of the story points of that user story to our velocity, or not?
I’ve never heard any agile or lean practitioner say YES. When the user story, feature or MMF is not completely 100% done, it has no value to the company. Now there are of course exceptions. For instance when you’re doing a spike, finishing 90% may actually bring substantial value to the company.
My personal opinion is, NO we don’t get any points, but for another reason. When you only count the story points of user stories that are completely done, problems in the team’s process will be visible in the velocity. For instance, a common cause is testing. Most teams don’t struggle with getting the development of a user story to done, it’s getting it trough functional testing.
Not receiving any points will have a shock effect, it will be clear to everyone that we have a problem and actions must be taken. An ideal message to take with you to the retrospective.
On the contrary, what if we would count 90% of the story points? The shock effect would not be there, and we might not search for improvements as hard as we could.
So why not keep it simple? If it’s done, we count the points. If it’s not done, we get nothing. Maybe difficult to accept at first, but it will pay off immediately if there’s something going on.