Fully Distributed Scrum: Linear Scalability of Production Between SFO and India
Download session PDF
Add to calendar
, Guido Schoonheim
The authors previously showed Scrum teams using XP practices achieved distributed velocity equal to local velocity with multiple distributed teams. Local velocity equaled distributed velocity and production increased linearly as teams scaled up to over 50 developers. Here we show a similar pattern for extreme time zone differences between San Francisco and India. Local velocity was established at five times industry average waterfall velocity. When team members were added in india, production scaled linearly. Detailed data on team process and performance will help others achieve the same goal.
Last year we had standing room only in our paper on distributed Scrum. We used 30 minutes for presenting the distributed team organization, practices, and detailed data on velocity and quality. We then had 15 minutes for discussion of details on team formation and overcoming distributed problems. We think there is significant interest is the distributed problem, particularly with extreme time zone differences as between Silicon Valley and India. This paper directly addresses those problems with detailed data on performance of both the team and the product in the marketplace.
- How to set up a fully distributed team
- How to measure team performance - velocity and quality
- How to assure local velocity is maintained when team distributes across time zones
- How to deal with cultural differences
- How to deal with extreme timezone differences