1

I'm looking for information and sources showing how typical length of iteration has changed over time.

Currently we see 2-week iteration used most often. I've seen some values between 1 and 2 months. Anyone can confirm or negate possibly redirecting to an information source?

Clarification: What I'm trying to get is information how typical iteration length looked like e.g. in 2003 or 2004 or something, possibly with some source. I generally know how it looks like these days and I know there isn't one standard size for every team/project.

flag
Are you looking for a trend line? Early publications said 30 days. Many folks now say 2 weeks. Are you saying that there's a correlation between publication date and recommended sprint duration? Or are you just asking if different people have different suggestions? I don't follow the question. – S.Lott Jan 25 at 23:08
No, not a trend line. Rather a source which says, the same as we are saying now "it's 2 weeks", it was usually 30 days in 2002 or something like this. Can you point 'early publications' you mention? – pawelbrodzinski Jan 26 at 6:30
1 
 Pawel, The Ken Schwaber books (published as recently as 2007) say 1 month is optimal. I think it was the hybrdising with XP that drove the push to 2 weeks (a compromise between the two agile superpowers?) – Craig Brown Jan 28 at 14:11

4 Answers

1

See "The Mystery of the Shrinking Story", by Jeff Patton here http://www.agileproductdesign.com/blog/the_shrinking_story.html

link|flag
0

FWIW I reckon the answer to what length should we have depends on your external stakeholders as much as internal.

And the next step in the shrinking iteration? Kanban, right?

link|flag
No. Time-boxing is very good when it makes sense. Anyway I am just looking for sources here. I don't try to make a point for or against any method. – pawelbrodzinski Jan 28 at 16:48
0

From my observations of sprints on software development projects in a corporation, the best length of a sprint is one months. I believe this is what all books recommend and my experiences agrees with that. Two main reasons against changing the lenght: 1) Making sprints shorter than a month puts too much overhead on the team. Both the sprint planning meeting and the sprint review meeting take significant time to prepare and to go over with all stakeholders. In the end, less time left to do the actual work. 2) Making sprints longer than a month makes upper management nervous about the project. From my observation, sponsors and key stakeholders would like to review the progress once a month. Otherwise they start to worry about the progress and pressure the team. To summary, 1 month is the best length of a spring in my experience.

link|flag
Actually it wasn't a question about the best iteration length. I added clarification. – pawelbrodzinski Jan 27 at 9:37
I'd disagree, shorter sprints result in a faster feedback cycle. As for overheads, your planning, retrospectives etc should all be proportionate, shorter sprints mean less to plan so shorter planning sessions. – Scrimmers Jan 27 at 15:46
0

Here's what I've seen in speaking with many of our clients.

  • Development projects generally run on 1 - 2 week sprints.

  • Projects related to operations/keeping the lights-on run in 3 week to 1 month sprints.

  • Projects related to achieving longer term, strategic goals run in 3 week to 3 month sprints (if you can call them that).

link|flag
What I'm trying to get is some historical data related to typical (development) projects. Added clarification to the question. – pawelbrodzinski Jan 27 at 9:34

Your Answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.