Extending Sprints
Nick Malik has posted parts of a conversation from an internal Microsoft agile discussion list. The question raised was “We are going to miss… do we stretch out the sprint?” The response was a big NO.
One of the responses suggested planning 2 weeks of “required” work and 4 weeks of “bonus” work for each sprint. That would give the team the opportunity to under-promise and over-deliver, but I cannot recall seeing that approach suggested in Scrum literature. It seems somewhat counter to the Scrum ideal. Has anyone tried it, or have any comments?
Comment
-
We have tried adding more that what is estimated to be achievable to sprint backlogs (on the insistence of management). Generally we have found that if you only add a reasonable amount more than recent velocity (say 10 – 20%) more, the team may be able to pull together to achieve that or come very close. If it is vastly more than recent velocity suggests (100 or 200% more) then no-one honestly believes that it is all achievable and the end is never in sight. You end up with a burn-down that stays high, a lack of predictability as to what gets done and what doesn’t (which can result in important expectations not being met), a weakening of discipline in estimating and committing as well as some degree of disappointment which can lower morale. Although I haven’t seen this first-hand so much, you also run the risk of sacrificing quality in an attempt to take on more work.
In my experience, it is much better to plan realistically and have the top of the uncommitted backlog sufficiently prepared for the team to be able to pull in additional top priority items towards the end of the sprint if they run out of work. There’s your bonus without creating false expectations.
-
Commenting is closed for this article.
« Venue Set for Fall 2007 Scrum Gathering : Agile2007 presentations reviewed at BarCampSydney »
21 August 2007, 13:26
You want to measure your team’s velocity. When you do that missing the sprint’s goal simply shows that you have put too much onto your plate. The velocity will decline and next time you plan, you should take on less work. After some iterations you will not miss again.