Here is a visual example of a team learning the value of a Scrum Sprint Burndown chart. To give some context to the images below - the team did a 3 day workshop on Scrum between sprint 1 & sprint 2. In this workshop we laid down some working agreements, one of which was they would use a physical Scrum task board with several major items. One item was the burndown chart derived from task hours estimated remaining each day. This team knew they were under committed for the sprint but wished to get started sprinting (a good choice rather than spend more time in sprint planning with so much uncertainty). They felt they could get to working on about 5 high priority stories and plan to replan (pull into the sprint more stories) later in the week. They assumed that learning about the uncertainty was better than crystal ball gazing.
The team started with 64 units (estimated task hours) of work. On the fourth day they decided to bring in more stories with estimated 24 units of work. Then on the fifth day they agreed to pull in an additional 34 units of work. This is their original burndown for the sprint.
What does this burndown tell us? It tells us that the team is not going to finish the sprint. It is only when one stops to read the fine print (the annotations) that one may gleam that there is more to the picture than is being visualized. Asking the team what they could do to make this chart "tell the whole truth", they decided to redraw the burndown.
Now this 2nd generation chart is telling a more complete story. It shows that on day four the team recognized it was finishing the committed work very early (as assumed - and then proven by tracking). So they pulled in some work on that day from the prioritized backlog. They were dealing with stories that had dependencies (not quite at the INVEST model standard - yet). But by doing work on the dependent stories the understanding of the team was growing. They pulled in 24 units of work (drawn below the zero datum on day 4). The very next day the team was in a similar position and pulled in an additional 34 units of work (units were derived from task estimates - not story points). The additional 34 units are drawn below the 24 datum line (at -58).
Will we get to DONE?
On day 8 the very quick progress slowed. This was not evident in the low fidelity charts and some team members were debating about the proper location of the top of day 8s bar on the graph. The solution to this "impediment" is to use graph paper. This allows for a more precise view and projection. Projecting the trend is the purpose of the burndown chart. We want to answer the question: Will we get to DONE?
At sprint 2 - I am much less concerned with the team getting all the work pulled in to the sprint Done, and much more concerned that they learn to use the tools and techniques that allow them to become self-managing. Next sprint they will have a better idea of the amount of work they can accomplish in a sprint (velocity). And if they have taken time to groom their backlog this sprint, next sprint's planning meeting will go much more smoothly. They may be able to fully plan the sprint and not need to pull in stories ad-hoc.
It is all part of learning to use the Scrum tools and techniques. Learning is exciting!
See Also: We have the best tools - why do we not use them?
The team started with 64 units (estimated task hours) of work. On the fourth day they decided to bring in more stories with estimated 24 units of work. Then on the fifth day they agreed to pull in an additional 34 units of work. This is their original burndown for the sprint.
What does this burndown tell us? It tells us that the team is not going to finish the sprint. It is only when one stops to read the fine print (the annotations) that one may gleam that there is more to the picture than is being visualized. Asking the team what they could do to make this chart "tell the whole truth", they decided to redraw the burndown.
Now this 2nd generation chart is telling a more complete story. It shows that on day four the team recognized it was finishing the committed work very early (as assumed - and then proven by tracking). So they pulled in some work on that day from the prioritized backlog. They were dealing with stories that had dependencies (not quite at the INVEST model standard - yet). But by doing work on the dependent stories the understanding of the team was growing. They pulled in 24 units of work (drawn below the zero datum on day 4). The very next day the team was in a similar position and pulled in an additional 34 units of work (units were derived from task estimates - not story points). The additional 34 units are drawn below the 24 datum line (at -58).
Will we get to DONE?
On day 8 the very quick progress slowed. This was not evident in the low fidelity charts and some team members were debating about the proper location of the top of day 8s bar on the graph. The solution to this "impediment" is to use graph paper. This allows for a more precise view and projection. Projecting the trend is the purpose of the burndown chart. We want to answer the question: Will we get to DONE?
At sprint 2 - I am much less concerned with the team getting all the work pulled in to the sprint Done, and much more concerned that they learn to use the tools and techniques that allow them to become self-managing. Next sprint they will have a better idea of the amount of work they can accomplish in a sprint (velocity). And if they have taken time to groom their backlog this sprint, next sprint's planning meeting will go much more smoothly. They may be able to fully plan the sprint and not need to pull in stories ad-hoc.
It is all part of learning to use the Scrum tools and techniques. Learning is exciting!
See Also: We have the best tools - why do we not use them?
Comments