The main target of scaling scrum is to provide at least one integrated increment for each sprint. The main scrum rule doesn’t change, and we only add some extra for working with multiple teams.


In this chapter, I made the same references to the Nexus Framework. Still, I avoid referencing it with low level detail because the purpose of this chapter is the Scaling Scrum and not only the Nexus Framework.


Starting with role and artifact, we have an integrated increment; it’s essential that all teams need to agree on the exact Definition of Done (DoD).

It’s important to know that even we have multiple teams, we have only one product. This means that:

  • We have only one product backlog, and all the team work on it;
  • We have only one product owner for all the team

As we know from the PSMI guide. If the product owner needs help, he can delegate part of its works, but a role like “Product owner proxy” doesn’t exist.

With this one product, we have multiple teams, where each person can be shared or dedicated to a single team (either developers or scrum master). But you need to remember that if a team member works on multiple teams, they lose their efficiency for the context switching problem.
So in details you can have:

  • Only one product owner for all the team;
  • One dedicated scrum master for each team, or one or more scrum master shared between the team (for example, one scrum master can serve multiple teams);
  • One or more Developers shared or not between teams (for example the super expert of some technology can be shared among the team);

All the scrum teams remain self-organized, so they need to coordinate. There isn’t a single coordinator.

Example: In image1, we have Developer 2.1 and Scrum Master 2  that work in Scrum Team 2 and Scrum Team 3.
Image1 – Scaling Scrum example


About events we have:

  • An high-level Sprint planning with a representative for each team. Next, we have a separate sprint planning for each group;
  • An high-level Daily Scrum representative for each team. Next, we have a different daily scrum for each group;
  • Only one integrated Sprint review where all the Scrum Teams present the integrated increment;
  • An high-level Sprint retrospective with a representative for each team. Next to it, we have a separate sprint review for each group.

About the sprint is better to have the same cadence (for example, a 30 days sprint) for all the teams, but if needed, we can have different timing for each sprint, but it’s essential that at each end of a sprint, at least one increment is released.

For the exam you need to remember that:

  • The target of scaling scrum is to provide at least one integrated increment for each sprint;
  • The main scrum rule doesn’t change;
  • All the teams need to agree on the exact definition of done;
  • Even with multiple teams, we have one product backlog and one product owner;
  • The scrum teams remain self-organized. There isn’t a coordinator;
  • Scrum master and Developers can be shared between teams.

References:

Share on: