Planning for Research and Customer defects
In my organization we are following SCRUM, but the most of the backlog stories are a mix of research stories (No prior work experience in that technology or functionality or domain) and customer defects (some time they are ad hoc but need to address with high priority, they also need some times more debugging effort)
we are failing every time in planning these stories leading huge variance and resources end up with working more hours to complete the story in the sprint- complete violation of scrum rules and spirit.
Our sprint length is 2 weeks.
Team is just giving estimated hours just for the sake of doing.
Please share your thoughts and suggestion and experience to handle and do proper planning with minimum variance.
How can we do backlog grooming in this case. what would be the correct way of planning
I appreciate your response.
> ... most of the backlog stories are a mix of research stories (No prior work
> experience in that technology or functionality or domain) and customer defects...
Would it be possible to alternate the Sprints, so that the first would address the stories while the second would address defects? This would separate the main concerns and hopefully introduce an element of stability. Since you are sprinting every 2 weeks, you should still be able to release an increment monthly.
It's always hard to estimate defects. In the most cases I suggest to dedicate % of capacity to work on defects. Is it possible in your case? If not, you might consider to use Kanban instead of Scrum.
In case of the lack of domain knowledge you can use a spike. Briefly it's a PBI to dig into a problem and the output should help to get knowledge in the area and provide more accurate estimation then.
Is it possible to temporarily change the length of sprints to 1 week? It would help to have often and shorter planning meetings.
Yes, having a refinement (formerly grooming) should help as well.