Forums

By posting to our forums you are agreeing to the Forum terms of use.
Question about optimal Sprint length
Last Post 21 Dec 2013 11:23 AM by Ian Mitchell. 6 Replies.
  •  
  •  
  •  
  •  
  •  
Sort:
PrevPrev NextNext
You are not authorized to post a reply.
Author Messages
cologne1986
New Member
New Member
Posts:6
cologne1986

--
19 Dec 2013 05:03 AM
    Hi,

    could you please help me with the following question?



    Optimal Sprint length?

    1.It is best to have Sprints of consistent length throughout a development effort.

    2. Sprint length is determined during Sprint Planning, and should be long enough to make sure the Development Team can deliver what is to be accomplished in the upcoming Sprint.

    3. All Sprints must be 1 month or less.

    4.The length of the Sprint should be proportional to the work that is done in between Sprints.

    5.Sprint length is determined during Sprint Planning, and should hold the time it will take to code the planned features in the upcoming Sprint, but does not include time for any testing.


    Thank you in advance!
    Ian Mitchell
    Advanced Member
    Advanced Member
    Posts:575
    Ian Mitchell

    --
    19 Dec 2013 05:29 AM
    1.It is best to have Sprints of consistent length throughout a development effort.

    The implementation of Scrum is better subjected to inspection and adaptation if the Sprints over which observations are compared are of consistent length. Therefore, this answer is true.

    2. Sprint length is determined during Sprint Planning, and should be long enough to make sure the Development Team can deliver what is to be accomplished in the upcoming Sprint.

    Sprint length is not a variable to be revised each Sprint, whether in Sprint Planning or during any other event. Moreover, Sprints do not flex in order to accommodate scope. Rather, it is the scope that is flexed in order to be accommodated within the available Sprint time-boxes.

    3. All Sprints must be 1 month or less.

    This is true. Sprints of greater than one month would mean feedback cycles becoming attenuated, and could result in the excessive batching of work. Sprints must indeed be no longer than 1 month.

    4.The length of the Sprint should be proportional to the work that is done in between Sprints.

    This is incorrect and it should be noted that Sprints must run consecutively with no work done between them.

    5.Sprint length is determined during Sprint Planning, and should hold the time it will take to code the planned features in the upcoming Sprint, but does not include time for any testing.

    Each of these assertions is incorrect. Sprint length is not determined during Sprint Planning, it is not planned around the upcoming Sprint features, and it should always be sufficient to allow the team to perform any testing required in order to meet their Definition of Done and to deliver a potentially releasable increment.
    cologne1986
    New Member
    New Member
    Posts:6
    cologne1986

    --
    19 Dec 2013 05:51 AM
    Thank you for quick response.
    I've also tends to the answers 1 and 3, but was not sure.
    Fabio Simeone
    New Member
    New Member
    Posts:14
    Fabio Simeone

    --
    19 Dec 2013 11:30 AM
    Hi Ian, I have a question. You said:

    "Sprint length is not a variable to be revised each Sprint, whether in Sprint Planning or during any other event. Moreover, Sprints do not flex in order to accommodate scope. Rather, it is the scope that is flexed in order to be accommodated within the available Sprint time-boxes. "

    So when do the Team decides the length? Let's pretend a new project is starting next week and the Team is already formed. How do they come up with the optimun length?
    Ian Mitchell
    Advanced Member
    Advanced Member
    Posts:575
    Ian Mitchell

    --
    19 Dec 2013 12:45 PM
    The Scrum Team would determine the optimum length by considering: organizational standards (e.g. an established release cadence); the ability of the Product Owner to phrase Sprint Goals within the length of the timebox and to release the value of each increment produced; and the Development Team's ability to forecast and deliver value within the given timebox.

    Scrum does not prescribe a time or event for making this determination. It could be during the *very first* Sprint Planning session. If sprint length absolutely had to change at some future point, this would be investigated and actioned during a Sprint Retrospective.
    Olivier Ledru
    New Member
    New Member
    Posts:31
    Olivier Ledru

    --
    21 Dec 2013 06:24 AM
    And what about the maximum risk (=> max Sprint length) tolerance ?
    Don't you think there is a correlation between risk of a project and the length of the Sprint ?
    Ian Mitchell
    Advanced Member
    Advanced Member
    Posts:575
    Ian Mitchell

    --
    21 Dec 2013 11:23 AM
    > Don't you think there is a correlation between
    > risk of a project and the length of the Sprint ?

    I think that risk-opportunity should inform organizational standards, the elicitation of sprint goals, and the forecasts made. The risks and opportunities present in business, project, and technical environments are all important and they can sometimes be in tension. This is why each should be considered when determining the optimum sprint length.
    You are not authorized to post a reply.


    Feedback