Scrum teams perform product backlog refinement (PBR) to address product backlog items (PBI) and maintain refined product backlog. For multiple teams, have workshops by distributing teammates to different groups. The groups then solve all the PBIs irrespective of the team they came from. They leave the workshop with a widened backlog perspective and an improved sense of shared responsibilities. In this article at Scrum.org, Cesario Ramos shares 4 formats for improved multi-team product backlog refinement.
Product Backlog Refinement for Multiple Teams
Product backlog refinement involves several activities:
- Understanding the primary problems teams need solving
- Further division of larger PBIs among groups for better understanding
- Analyzing items to find out the learning scope and alignment to business goals
- Probing deep to discover precise customer requirements
- Create acceptance tests or user stories with Gherkin specifications, flow tables, etc. by comprehending items better
- Recognizing risks and features that need manual testing and log those in exploratory test charters
- Focusing on composing initial models of the solution
Refinement with Several Teams:
For multi-team product backlog refinement, all come together and work on PBIs as per workshop groups and not team. The benefits of these workshops are astounding:
- Increased Flexibility in PBI Distribution: As all teams have handled all PBIs, you can distribute items without waiting for the parent team to become free.
- Intuitive Knowledge of Dependencies: With a holistic understanding of the product, teams can predict probable dependencies and coordinate accordingly.
- Product-Level Status Visibility: Teams share a product-level velocity rather than a siloed velocity.
4 Formats for Multi-Team PBRs:
- Full Roulette: Divide teams into random groups that take up a PBI and refine it in their station. After 15 minutes, groups move to the next station and work on the PBI left by the previous group. This rotation continues until they have refined all the PBIs, irrespective of the method they chose for product backlog refinement. The results are good as every group needs to do a neat job for the next group.
- Partial Roulette: Instead of the entire group moving to the next station, one of the group members stays back. That individual passes on lessons learned by the previous group. You can start with this format and change to Full Roulette when everyone is settled in.
- Diverge and Merge: In this format, not only does a group member stay back at the station, another member rotates through all groups. Gathering all the learnings, the rotating member returns to the original group and share information. Discussions follow henceforth. The author had successful outcomes with this format while working with up to 35 people.
- Teach Back: Each group picks up a PBI and refines it. After the 15-minute timeline is over, group members teach others and clarify doubts. However, this format does not work well with bigger groups because of diversified interests.
To view the original article in full, visit the following link: https://www.scrum.org/resources/blog/multi-team-backlog-refinement