Loading ...
Sorry, an error occurred while loading the content.

Sprint Length

Expand Messages
  • Loren
    When I first became certified I had learned that a sprint was a 30 day cycle. Now I am hearing more and more people talking about shorter sprints: 3 week, 2
    Message 1 of 17 , Jan 10, 2009
    • 0 Attachment
      When I first became certified I had learned that a sprint was a 30 day
      cycle. Now I am hearing more and more people talking about shorter
      sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
      short? In your opinion what is the ideal sprint length? Is a 30 day
      cycle ideal because it gives you time to produce more functionality,
      more deliverables?
    • Roy Morien
      Let s think about s sprint and the activities in a sprint. A sprint has an administrative overhead. Could we state a proportion of sprint time for that
      Message 2 of 17 , Jan 10, 2009
      • 0 Attachment
        Let's think about s sprint and the activities in a sprint.
         
        A sprint has an 'administrative' overhead. Could we state a proportion of sprint time for that overhead, which includes sprint planning, end of sprint demo and acceptance etc. Maybe 10% of sprint time is ok, so if the actual time taken is greater than this, then the sprint should be lengthened .. but not the admin overhead time.
         
        User stories should be able to be completed within a sprint, so big User Stories and Epics should be broken down into smaller User Stories. If you find that those smaller User Stories are becoming rather trivial, just to fit into the sprint, then maybe the sprint length should be lengthened. The corollory to this might be if you must select an unduly large number of User Stories to fill a sprint. Either the User Stories are too small, or the sprint should be shortened.
         
        Sprints allow regular and frequent validation and verification of development activity in its various forms. If you think that this is not happening frequently enough, then perhaps the sprint length should be shortened.

        I think these are all rather imprecise notions, indicating that the 'perfect' length of a sprint depends on a few very important factors which are hard to quantify..
         
        Personally, I would adopt a learning approach to this (hmmm, is that 'agile thinking'? :))
         
        Try a week. If it works, and everyone sees value in that, then fine. Otherwise, try 2 weeks. If you are perfectly happy with 30 day sprints, which I assume to be 4 weeks, or even a calendar month, rather than 30 days precisely, then again, go for it.
         
        Your question is not about 'help me my 30 day sprints are not working' so i assume that you are content with the status quo. You have a choice therefore. Adopt the 'if it ain't broke, don't fix it' view, and stay with 30 day sprints. Or, you might think that a shorter sprint could be more effective, but you don't know that .... so try it. Or a 1 week sprint, perhaps.
         
        Regards,
        Roy Morien


        To: scrumdevelopment@yahoogroups.com
        From: loren.silverman@...
        Date: Sat, 10 Jan 2009 12:13:19 +0000
        Subject: [scrumdevelopment] Sprint Length


        When I first became certified I had learned that a sprint was a 30 day
        cycle. Now I am hearing more and more people talking about shorter
        sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
        short? In your opinion what is the ideal sprint length? Is a 30 day
        cycle ideal because it gives you time to produce more functionality,
        more deliverables?




        Find great deals on eBay Net yourself a bargain
      • Ron Jeffries
        Hello, Loren. On Saturday, January 10, 2009, at 7:13:19 AM, you ... Yes. A minute would be too short. Probably a day would be. ... I like one week. It s
        Message 3 of 17 , Jan 10, 2009
        • 0 Attachment
          Hello, Loren. On Saturday, January 10, 2009, at 7:13:19 AM, you
          wrote:

          > When I first became certified I had learned that a sprint was a 30 day
          > cycle. Now I am hearing more and more people talking about shorter
          > sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
          > short?

          Yes. A minute would be too short. Probably a day would be.

          > In your opinion what is the ideal sprint length?

          I like one week. It's difficult and many teams are not up to it,
          however.

          > Is a 30 day
          > cycle ideal because it gives you time to produce more functionality,
          > more deliverables?

          No, it is far from ideal, in my opinion, because it gives you more
          time to waste.

          Ron Jeffries
          www.XProgramming.com
          www.xprogramming.com/blog
          To be on the wire is life. The rest is waiting. --Karl Wallenda
        • Richard Lawrence
          I like one week sprints. As Ron said, it s difficult--it forces you to do difficult things like decomposing stories very small, working closely across roles,
          Message 4 of 17 , Jan 10, 2009
          • 0 Attachment
            I like one week sprints. As Ron said, it's difficult--it forces you to
            do difficult things like decomposing stories very small, working
            closely across roles, completing things every day, getting frequent
            Product Owner feedback, etc. But that's one of the things I like about
            it; it forces the team to learn good habits. Long sprints allow bad
            habits to form. The two I see most often are student syndrome, working
            slow at the beginning of a sprint and like crazy at the end, and
            mini-waterfalls, structuring each sprint like a waterfall project with
            design, build, test, etc.

            Roy is right that sprints have so-called administrative overhead. But
            in my experience, that overhead grows non-linearly as the sprint
            length increases. If a team takes an hour to plan a one week sprint,
            I'd expect them to take two and a quarter or two and a half hours to
            plan a two week sprint. That's because energy and focus are better in
            short meetings. One team I coached was spending over a day planning a
            30 day sprint but cut it to an hour when they switched to one week
            sprints.

            There's one other benefit to short sprints: I find that a team's
            comfort with agile and level of improvement are more correlated with
            the number of sprints they've done than the amount of time they've
            been using Scrum. With one week sprints it's possible to see many of
            the same improvements in three weeks that a team with 30 day sprints
            would see in twelve.

            Richard
            --
            Richard Lawrence
            Certified Scrum Coach
            Founder and Principal Consultant, Humanizing Work, LLC
            303-895-7688
            richard@...
            www.humanizingwork.com
            www.richardlawrence.info

            On Sat, Jan 10, 2009 at 5:13 AM, Loren <loren.silverman@...> wrote:
            > When I first became certified I had learned that a sprint was a 30 day
            > cycle. Now I am hearing more and more people talking about shorter
            > sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
            > short? In your opinion what is the ideal sprint length? Is a 30 day
            > cycle ideal because it gives you time to produce more functionality,
            > more deliverables?
            >
            >
          • Michael Yip
            Hi Richard, How do you define benefits gained between 1 week Sprint and greater than 1 week Sprints? What is your definition of Done for a 1 week Sprint or
            Message 5 of 17 , Jan 10, 2009
            • 0 Attachment
              Hi Richard,

              How do you define benefits gained between 1 week Sprint and greater than 1 week Sprints? What is your definition of "Done" for a 1 week Sprint or is there one? Have you seen the quality of work (defects escaped between Sprints and defects escaped in the field) between a 1 week Sprint and greater than 1 week Sprint with similar project complexity and team capability? Do you have correlation between these defects base on severity to the quality of stories produced between 1 week Sprint and greater than 1 week Sprints? Are there any long term impact to the morale and quality of work for teams consistently doing 1 week Sprint versus greater than 1 week Sprints?

              Regards,
              Michael


              --- On Sat, 1/10/09, Richard Lawrence <rslawrence@...> wrote:

              From: Richard Lawrence <rslawrence@...>
              Subject: Re: [scrumdevelopment] Sprint Length
              To: scrumdevelopment@yahoogroups.com
              Date: Saturday, January 10, 2009, 11:38 AM

              I like one week sprints. As Ron said, it's difficult--it forces you to
              do difficult things like decomposing stories very small, working
              closely across roles, completing things every day, getting frequent
              Product Owner feedback, etc. But that's one of the things I like about
              it; it forces the team to learn good habits. Long sprints allow bad
              habits to form. The two I see most often are student syndrome, working
              slow at the beginning of a sprint and like crazy at the end, and
              mini-waterfalls, structuring each sprint like a waterfall project with
              design, build, test, etc.

              Roy is right that sprints have so-called administrative overhead. But
              in my experience, that overhead grows non-linearly as the sprint
              length increases. If a team takes an hour to plan a one week sprint,
              I'd expect them to take two and a quarter or two and a half hours to
              plan a two week sprint. That's because energy and focus are better in
              short meetings. One team I coached was spending over a day planning a
              30 day sprint but cut it to an hour when they switched to one week
              sprints.

              There's one other benefit to short sprints: I find that a team's
              comfort with agile and level of improvement are more correlated with
              the number of sprints they've done than the amount of time they've
              been using Scrum. With one week sprints it's possible to see many of
              the same improvements in three weeks that a team with 30 day sprints
              would see in twelve.

              Richard
              --
              Richard Lawrence
              Certified Scrum Coach
              Founder and Principal Consultant, Humanizing Work, LLC
              303-895-7688
              richard@humanizingw ork.com
              www.humanizingwork. com
              www.richardlawrence .info

              On Sat, Jan 10, 2009 at 5:13 AM, Loren <loren.silverman@ gmail.com> wrote:
              > When I first became certified I had learned that a sprint was a 30 day
              > cycle. Now I am hearing more and more people talking about shorter
              > sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
              > short? In your opinion what is the ideal sprint length? Is a 30 day
              > cycle ideal because it gives you time to produce more functionality,
              > more deliverables?
              >
              >

            • Cenk Çivici
              Thats interesting. Why do Scrum Trainers teach people to do 30 day cycles?
              Message 6 of 17 , Jan 10, 2009
              • 0 Attachment
                Thats interesting. Why do Scrum Trainers teach people to do 30 day cycles?

                On Sat, Jan 10, 2009 at 2:13 PM, Loren <loren.silverman@...> wrote:
                > When I first became certified I had learned that a sprint was a 30 day
                > cycle. Now I am hearing more and more people talking about shorter
                > sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
                > short? In your opinion what is the ideal sprint length? Is a 30 day
                > cycle ideal because it gives you time to produce more functionality,
                > more deliverables?
                >
                >
              • Richard Lawrence
                That s a lot of questions, but I ll try. :) See inline... ... In general, I see more of the behaviors I like with teams using shorter sprints than those using
                Message 7 of 17 , Jan 10, 2009
                • 0 Attachment
                  That's a lot of questions, but I'll try. :) See inline...

                  On Sat, Jan 10, 2009 at 11:25 AM, Michael Yip <myip11@...> wrote:
                  > Hi Richard,
                  >
                  > How do you define benefits gained between 1 week Sprint and greater than 1
                  > week Sprints?

                  In general, I see more of the behaviors I like with teams using
                  shorter sprints than those using longer sprints. These include:
                  * Smaller stories
                  * More interaction across roles
                  * A greater willingness to try ATDD and an easier time adopting it
                  * A tendency to show real progress every day
                  * An even pace across the sprint instead of a crazy push at the end of
                  each sprint
                  * Faster learning and adaptation

                  It's not a perfect correlation, of course. I've seen teams with 2-3
                  week sprints that do all the things I like to see. And I've seen teams
                  with 1 week sprints continue to struggle. But I find that short
                  sprints make certain bad habits impossible to get away with and
                  certain good habits more attractive to learn.

                  > What is your definition of "Done" for a 1 week Sprint or is there one?

                  This is important. Every product backlog item must get to DONE in the
                  sprint. In a 1 week sprint, there should just be fewer and/or smaller
                  items than in a longer sprint.

                  > Have you seen the quality of work (defects escaped between Sprints and defects escaped in the
                  > field) between a 1 week Sprint and greater than 1 week Sprint with similar project complexity and
                  > team capability? Do you have correlation between these defects base on severity
                  > to the quality of stories produced between 1 week Sprint and greater than 1
                  > week Sprints?

                  In general, I've seen better quality from teams with short sprints.
                  Not directly because of the sprint length, but because other other
                  quality-related habits a short sprint seems to promote. (I've worked
                  with only a few dozen teams and no two are exactly the same, so this
                  is just anecdotal, of course.)

                  > Are there any long term impact to the morale and quality of
                  > work for teams consistently doing 1 week Sprint versus greater than 1 week
                  > Sprints?

                  Sustainable pace seems to be easier to achieve with short sprints
                  because it precludes student syndrome and allows the team to establish
                  an empirical velocity very quickly. But I can see this going the other
                  way if the team overcommits each week (and have heard reports to this
                  effect). That, though, is not a problem of sprint length but a problem
                  of committing appropriately.

                  Richard

                  >
                  > Regards,
                  > Michael
                  >
                  >
                  > --- On Sat, 1/10/09, Richard Lawrence <rslawrence@...> wrote:
                  >
                  > From: Richard Lawrence <rslawrence@...>
                  > Subject: Re: [scrumdevelopment] Sprint Length
                  > To: scrumdevelopment@yahoogroups.com
                  > Date: Saturday, January 10, 2009, 11:38 AM
                  >
                  > I like one week sprints. As Ron said, it's difficult--it forces you to
                  > do difficult things like decomposing stories very small, working
                  > closely across roles, completing things every day, getting frequent
                  > Product Owner feedback, etc. But that's one of the things I like about
                  > it; it forces the team to learn good habits. Long sprints allow bad
                  > habits to form. The two I see most often are student syndrome, working
                  > slow at the beginning of a sprint and like crazy at the end, and
                  > mini-waterfalls, structuring each sprint like a waterfall project with
                  > design, build, test, etc.
                  >
                  > Roy is right that sprints have so-called administrative overhead. But
                  > in my experience, that overhead grows non-linearly as the sprint
                  > length increases. If a team takes an hour to plan a one week sprint,
                  > I'd expect them to take two and a quarter or two and a half hours to
                  > plan a two week sprint. That's because energy and focus are better in
                  > short meetings. One team I coached was spending over a day planning a
                  > 30 day sprint but cut it to an hour when they switched to one week
                  > sprints.
                  >
                  > There's one other benefit to short sprints: I find that a team's
                  > comfort with agile and level of improvement are more correlated with
                  > the number of sprints they've done than the amount of time they've
                  > been using Scrum. With one week sprints it's possible to see many of
                  > the same improvements in three weeks that a team with 30 day sprints
                  > would see in twelve.
                  >
                  > Richard
                  > --
                  > Richard Lawrence
                  > Certified Scrum Coach
                  > Founder and Principal Consultant, Humanizing Work, LLC
                  > 303-895-7688
                  > richard@humanizingw ork.com
                  > www.humanizingwork. com
                  > www.richardlawrence .info
                  >
                  > On Sat, Jan 10, 2009 at 5:13 AM, Loren <loren.silverman@ gmail.com> wrote:
                  >> When I first became certified I had learned that a sprint was a 30 day
                  >> cycle. Now I am hearing more and more people talking about shorter
                  >> sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
                  >> short? In your opinion what is the ideal sprint length? Is a 30 day
                  >> cycle ideal because it gives you time to produce more functionality,
                  >> more deliverables?
                  >>
                  >>
                  >
                  >
                • Richard Lawrence
                  Probably because that s what the original Scrum book prescribed. Most teams I come across today are doing 2-3 week sprints, and all the trainers I know
                  Message 8 of 17 , Jan 10, 2009
                  • 0 Attachment
                    Probably because that's what the original Scrum book prescribed.

                    Most teams I come across today are doing 2-3 week sprints, and all the
                    trainers I know personally teach that.

                    Richard

                    On Sat, Jan 10, 2009 at 11:51 AM, Cenk Çivici <cenk.civici@...> wrote:
                    > Thats interesting. Why do Scrum Trainers teach people to do 30 day cycles?
                    >
                    > On Sat, Jan 10, 2009 at 2:13 PM, Loren <loren.silverman@...> wrote:
                    >> When I first became certified I had learned that a sprint was a 30 day
                    >> cycle. Now I am hearing more and more people talking about shorter
                    >> sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
                    >> short? In your opinion what is the ideal sprint length? Is a 30 day
                    >> cycle ideal because it gives you time to produce more functionality,
                    >> more deliverables?
                    >>
                    >>
                    >
                  • William Wake
                    ... It s hard to know what everybody does. There are certainly those who talk about other sprint lengths. But I ve got to think that even if they talk about
                    Message 9 of 17 , Jan 10, 2009
                    • 0 Attachment
                      On Sat, Jan 10, 2009 at 1:51 PM, Cenk Çivici <cenk.civici@...> wrote:
                      > Thats interesting. Why do Scrum Trainers teach people to do 30 day cycles?

                      It's hard to know what everybody does. There are certainly those who
                      talk about other sprint lengths. But I've got to think that even if
                      they talk about sprint length "by the book" that they still talk about
                      "inspect and adapt":)

                      --Bill

                      --
                      Bill Wake William.Wake@... www.xp123.com
                    • Mitch Lacey
                      Every trainer I know today says that sprints should be no more than 30 days. There are many factors to consider in determining sprint length. Factors like: ·
                      Message 10 of 17 , Jan 10, 2009
                      • 0 Attachment

                        Every trainer I know today says that sprints should be no more than 30 days.

                         

                        There are many factors to consider in determining sprint length. Factors like:

                        ·         How much feedback the customer/stakeholders can handle

                        ·         How much time the customer/stakeholders can devote

                        ·         How educated in Scrum the team is

                        ·         How educated in Scrum the customer/stakeholders are

                        ·         Regulatory issues and compliances

                        ·         Existing cultural barriers in the company or with the stakeholder/customer group

                        ·         How good the core team is in decomposing the work

                        ·         How the Scrum team has set up their team to deliver – meaning the way you approach technical aspects in a 1-week sprint are significantly different than that of a 4-week sprint

                        o   Factors like automated acceptance testing, TDD, automated releases and the like

                         

                        I know of teams at Microsoft who build and manage online services. One team in particular is able to plan on Monday morning and release to production on Friday at 4:00pm. They never work weekends. It took them nearly two years to be able to do this – but they started out at 1-week sprints and stuck with it – and it was painful.

                         

                        From: scrumdevelopment@yahoogroups.com [mailto:scrumdevelopment@yahoogroups.com] On Behalf Of Richard Lawrence
                        Sent: Saturday, January 10, 2009 12:19 PM
                        To: scrumdevelopment@yahoogroups.com
                        Subject: Re: [scrumdevelopment] Sprint Length

                         

                        Probably because that's what the original Scrum book prescribed.

                        Most teams I come across today are doing 2-3 week sprints, and all the
                        trainers I know personally teach that.

                        Richard

                        On Sat, Jan 10, 2009 at 11:51 AM, Cenk Çivici <cenk.civici@...> wrote:

                        > Thats interesting. Why do Scrum Trainers teach people to do 30 day cycles?
                        >
                        > On Sat, Jan 10, 2009 at 2:13 PM, Loren <
                        href="mailto:loren.silverman%40gmail.com">loren.silverman@...> wrote:
                        >> When I first became certified I had learned that a sprint was a 30 day
                        >> cycle. Now I am hearing more and more people talking about shorter
                        >> sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
                        >> short? In your opinion what is the ideal sprint length? Is a 30 day
                        >> cycle ideal because it gives you time to produce more functionality,
                        >> more deliverables?
                        >>
                        >>
                        >

                      • Ken Schwaber
                        For most products, 30 days (or one calendar month) is a good balance between time needed to pull something potentially shippable together, and the risk that
                        Message 11 of 17 , Jan 10, 2009
                        • 0 Attachment
                          For most products, 30 days (or one calendar month) is a good balance between time needed to pull something potentially shippable together, and the risk that can be tolerated by not knowing where you are for greater than 30 days/1 month. Longer Sprints tend to require documentation for everyone to remember what is happen, increase the risk by going outside the normal business cycle of one month. and provide enough time for people to go back to waterfall thinking. Shorter Sprints (2 weeks, 1 week) are done when more predictability and risk management is desired, but you pay for it with a high just-in-time planning overhead. If there are more than one Scrum team, their sprint must all start and end the same time. They must integrate all of their work into one integrated increment.
                          Ken

                          On Jan 10, 2009, at 1:51 PM, Cenk Çivici wrote:

                          Thats interesting. Why do Scrum Trainers teach people to do 30 day cycles?

                          On Sat, Jan 10, 2009 at 2:13 PM, Loren <loren.silverman@ gmail.com> wrote:
                          > When I first became certified I had learned that a sprint was a 30 day
                          > cycle. Now I am hearing more and more people talking about shorter
                          > sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
                          > short? In your opinion what is the ideal sprint length? Is a 30 day
                          > cycle ideal because it gives you time to produce more functionality,
                          > more deliverables?
                          >
                          > 


                        • Ron Jeffries
                          Hello, Cenk. On Saturday, January 10, 2009, at 1:51:23 PM, you ... Many do not. Ken even mentions two weeks as having merit. Ron Jeffries www.XProgramming.com
                          Message 12 of 17 , Jan 10, 2009
                          • 0 Attachment
                            Hello, Cenk. On Saturday, January 10, 2009, at 1:51:23 PM, you
                            wrote:

                            > Thats interesting. Why do Scrum Trainers teach people to do 30 day cycles?

                            Many do not. Ken even mentions two weeks as having merit.

                            Ron Jeffries
                            www.XProgramming.com
                            www.xprogramming.com/blog
                            It's easy to have a complicated idea.
                            It's very very hard to have a simple idea.
                            -- Carver Mead
                          • Hugo Palma
                            I d like to add another very important factor which is the project length estimate. For example, if you did an rough project length estimate and got 2 months
                            Message 13 of 17 , Jan 11, 2009
                            • 0 Attachment
                              I'd like to add another very important factor which is the project length estimate. For example, if you did an rough project length estimate and got 2 months then definitely 30 days sprints would not be a good fit here.



                              On Sat, Jan 10, 2009 at 8:58 PM, Mitch Lacey <mglacey@...> wrote:

                              Every trainer I know today says that sprints should be no more than 30 days.

                               

                              There are many factors to consider in determining sprint length. Factors like:

                              ·         How much feedback the customer/stakeholders can handle

                              ·         How much time the customer/stakeholders can devote

                              ·         How educated in Scrum the team is

                              ·         How educated in Scrum the customer/stakeholders are

                              ·         Regulatory issues and compliances

                              ·         Existing cultural barriers in the company or with the stakeholder/customer group

                              ·         How good the core team is in decomposing the work

                              ·         How the Scrum team has set up their team to deliver – meaning the way you approach technical aspects in a 1-week sprint are significantly different than that of a 4-week sprint

                              o   Factors like automated acceptance testing, TDD, automated releases and the like

                               

                              I know of teams at Microsoft who build and manage online services. One team in particular is able to plan on Monday morning and release to production on Friday at 4:00pm. They never work weekends. It took them nearly two years to be able to do this – but they started out at 1-week sprints and stuck with it – and it was painful.

                               

                              From: scrumdevelopment@yahoogroups.com [mailto:scrumdevelopment@yahoogroups.com] On Behalf Of Richard Lawrence
                              Sent: Saturday, January 10, 2009 12:19 PM
                              To: scrumdevelopment@yahoogroups.com


                              Subject: Re: [scrumdevelopment] Sprint Length

                               

                              Probably because that's what the original Scrum book prescribed.



                              Most teams I come across today are doing 2-3 week sprints, and all the
                              trainers I know personally teach that.

                              Richard

                              On Sat, Jan 10, 2009 at 11:51 AM, Cenk Çivici <cenk.civici@...> wrote:
                              > Thats interesting. Why do Scrum Trainers teach people to do 30 day cycles?
                              >
                              > On Sat, Jan 10, 2009 at 2:13 PM, Loren <loren.silverman@...> wrote:
                              >> When I first became certified I had learned that a sprint was a 30 day
                              >> cycle. Now I am hearing more and more people talking about shorter
                              >> sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
                              >> short? In your opinion what is the ideal sprint length? Is a 30 day
                              >> cycle ideal because it gives you time to produce more functionality,
                              >> more deliverables?
                              >>
                              >>
                              >



                            • Cenk Çivici
                              Tell me if this is wrong. I always recommend beginning with shorter cycles for more frequent feedback especially when the team is new to Agile. 1 week
                              Message 14 of 17 , Jan 11, 2009
                              • 0 Attachment
                                Tell me if this is wrong. I always recommend beginning with shorter
                                cycles for more frequent feedback especially when the team is new to
                                Agile. 1 week iteration length for instance.

                                If this puts pressure on the dev team , then change it to 2 weeks or
                                3-4 weeks. I dont recommend the other way around which is beginning
                                from 4 weeks and shortening it to get more frequent feedback as you go
                                along.

                                On Sun, Jan 11, 2009 at 1:05 AM, Ron Jeffries <ronjeffries@...> wrote:
                                > Hello, Cenk. On Saturday, January 10, 2009, at 1:51:23 PM, you
                                > wrote:
                                >
                                >> Thats interesting. Why do Scrum Trainers teach people to do 30 day cycles?
                                >
                                > Many do not. Ken even mentions two weeks as having merit.
                                >
                                > Ron Jeffries
                                > www.XProgramming.com
                                > www.xprogramming.com/blog
                                > It's easy to have a complicated idea.
                                > It's very very hard to have a simple idea.
                                > -- Carver Mead
                                >
                                >
                              • Ron Jeffries
                                Hello, Cenk. On Sunday, January 11, 2009, at 8:30:33 AM, you ... There probably is no wrong . I prefer short iterations. I think the sharpen both the dev
                                Message 15 of 17 , Jan 11, 2009
                                • 0 Attachment
                                  Hello, Cenk. On Sunday, January 11, 2009, at 8:30:33 AM, you
                                  wrote:

                                  > Tell me if this is wrong. I always recommend beginning with shorter
                                  > cycles for more frequent feedback especially when the team is new to
                                  > Agile. 1 week iteration length for instance.

                                  > If this puts pressure on the dev team , then change it to 2 weeks or
                                  > 3-4 weeks. I dont recommend the other way around which is beginning
                                  > from 4 weeks and shortening it to get more frequent feedback as you go
                                  > along.

                                  There probably is no "wrong". I prefer short iterations. I think the
                                  sharpen both the dev team's ability to get done, and the PO's
                                  ability to think in terms of small useful bits to get done.

                                  I feel that "context-driven" notions like "too hard" or "puts
                                  pressure" are dangerously wimpy. On the other hand you don't see me
                                  do the 10 pushup challenge, do you? So teams have to do as they see
                                  fit ... but they also have to be careful not to wind up fat and old,
                                  like me.

                                  Ron Jeffries
                                  www.XProgramming.com
                                  www.xprogramming.com/blog
                                  And thirdly, the Code is more what you'd call guidelines than actual rules.
                                  -- Barbossa
                                • Cat Schwamm
                                  ... Loren, I agree with what Roy said, if it ain t broke don t fix it. Your teams should be working at a pace with which they are comfortable. My team, for
                                  Message 16 of 17 , Jan 12, 2009
                                  • 0 Attachment
                                    --- In scrumdevelopment@yahoogroups.com, "Loren" <loren.silverman@...>
                                    wrote:
                                    >
                                    > When I first became certified I had learned that a sprint was a 30 day
                                    > cycle. Now I am hearing more and more people talking about shorter
                                    > sprints: 3 week, 2 week, and even 1 week. Can a Sprint cycle be too
                                    > short? In your opinion what is the ideal sprint length? Is a 30 day
                                    > cycle ideal because it gives you time to produce more functionality,
                                    > more deliverables?
                                    >
                                    Loren,

                                    I agree with what Roy said, if it ain't broke don't fix it. Your
                                    teams should be working at a pace with which they are comfortable. My
                                    team, for example, works on a 2-month release cycle with 4 2-week
                                    sprints. The two weeks is great because we can get semi-large items
                                    taken care of (with enough time to fix any bugs) and our testers have
                                    plenty of time to make sure that everything is clean and according to
                                    spec.

                                    However, it is all based on the size and proficiency of a team; if you
                                    set them to a shorter sprint and they are unable to finish it, or do
                                    not release quality software, then you are only doing harm. I do
                                    occasionally see the "student syndrome" that Richard mentioned, so any
                                    longer than two weeks for me is just painful.

                                    And yes, one minute is -probably- too short of a sprint :] I'd love to
                                    get my team on one-week sprints, but they'd have my head on a stake
                                    before we could even do daily stand-up.

                                    Cheers,
                                    Cat
                                  • Richard Lawrence
                                    ... I wouldn t be too quick to make that judgment. If they are unable to finish or to reach the desired quality once or twice, that s good information. You may
                                    Message 17 of 17 , Jan 13, 2009
                                    • 0 Attachment
                                      On Mon, Jan 12, 2009 at 1:03 PM, Cat Schwamm <cat.schwamm@...> wrote:
                                      >
                                      > However, it is all based on the size and proficiency of a team; if you
                                      > set them to a shorter sprint and they are unable to finish it, or do
                                      > not release quality software, then you are only doing harm.

                                      I wouldn't be too quick to make that judgment. If they are unable to
                                      finish or to reach the desired quality once or twice, that's good
                                      information. You may have revealed problems that were previously
                                      hidden by a longer sprint length. Before giving up on the short
                                      sprints, use the retrospective to dig into why the team struggles to
                                      get done in a week and what they could change to be able to do it. I
                                      wouldn't leave a team struggling with short sprints for more than one
                                      or two, but I wouldn't start by assuming that it's not possible for
                                      that team to deliver in a week.

                                      Richard
                                    Your message has been successfully submitted and would be delivered to recipients shortly.