Estimative problems
So, my team asked me today if it's worth to document the wrong estimation on taks.
For example, If the dev estimate 2 hours for a task, but it took him 8 hours to finish it, this 6 plus hours affects the development estimated hour.. So they asked me if they could document it in the kanbam! I think they shouldn't and they should estimate better those tasks, but I wanted to get skilled opinion, what do you think?
I think they shouldn't and they should estimate better those tasks
Perhaps they could do so more efficiently if they captured relevant data and metrics. That sounds like what they are trying to do.
For each task they may choose to capture the original estimate, the current estimate of work remaining, and when it is finished the actual time it took.
it's worth to document the wrong estimation on tasks
I consider this a good inspect-and-adapt exercise and it's pretty simple to capture.
- It helps to identify similar work and rise a flag of caution at the moment of estimation.
- It helps to understand what tasks are often miss-estimated and more importantly: why.
- Far from being too obsesive with accuracy (and failing on it is perfectly fine), the more accurate a team is on estimations, the more 'sprint predictive' is, and less time is wasted within the initial forecast.
- It helps to detect areas of improvement, impediments not raised and/or lack of skills on specific things.
I think they shouldn't and they should estimate better those tasks.
Perhaps estimating better needs to go through a process of learning from inspecting their convictions first.
+1 on Juan's comment.
Definitely document this information for the point of inspection and adaptation; Scrum is all about empiricism after all right?