It is not over till it's over. And for that last mile to get past the finishing line, you may need to consider different strategy than the one for start and mid way.
Q: Are you kidding? Changing the ongoing processes, discarding honorable agile model and switching to kanban, and that too when we are so close to our release?
A: Hmmm, that's right.
Strategy which provided a good start and a smooth mid-way may or may not be the best for end.
Agile is awesome, quick releases and early user feedback really helps in planning and prioritising the way forward. But there are cases when kanban can be useful as well and could improve team's productivity and product quality.
Why kanban for the last mile?
In agile model, sprints are more of a step jump, let's say 1 sprint takes you takes you 50 steps forward (that's your sprint velocity). But in the last mile you are not sure if you need to walk just 20 steps or 70 steps, and regression testing may surprise by adding 30 more steps or maybe adding nothing more. Hence considering these variables kanban suits well to get par the finishing line and release your software.
But, wait it is still not over. What If the release wasn't received well and you need to patch it up. In this case you have a clear vision that patch will consist of some specific issues only, say 10 issues. Then isn't it better to put those 10 tickets on board, move them to done and release the patch. Seems much simpler approach the squeezing in those 10 issues in a sprint.
Another point to note here is that for these fixes and finishing that last mile, we don't need to go through all the planning and gromming phase of sprint, consider various technical solutions and decide on which one to pick. Hence, better to skip scrum ceremonies and get started rightaway with a small board with patch issues on it.
To conclude, for continuous and consistent delivery agile works well, and for important milestone releases kanban can deliver the final blow.
- Ayush 🙂