Product Backlog maintenance
Although it is not a formal Scrum event, in practice most teams carry it out periodically during the sprint, setting aside dedicated time for it. It is an ongoing activity rather than a calendar event: the Scrum Guide describes it as the act of breaking down and further defining Product Backlog items, without prescribing meetings or attendees.
Characteristics
- It is an activity parallel to the sprint: it does not interrupt the development in progress, but is carried out alongside it.
- It is led by the Product Owner, although the team actively participates in estimation and in the analysis of dependencies.
- The effort it consumes varies with the product and the team, and should be measured in each case; as a guideline, it usually ranges between 5% and 10% of the sprint's capacity.
- It has a specific goal per session: the aim is not to review the entire Product Backlog, but to make sure that the candidate stories for the next sprint have the necessary level of detail.
Common activities
- Adding new user stories based on customer feedback or newly identified needs.
- Splitting stories that are too large (epics) into smaller, more manageable stories.
- Rethinking or rewriting already defined stories that have become outdated.
- Estimating stories that have not yet been sized.
- Reordering the Product Backlog in response to changes in business priorities.
Role for the team and the Product Owner
Refinement serves a dual purpose:
- For the Product Owner, it is the opportunity to get feedback from the team on the size and complexity of the stories they are adding, before committing to include them in a sprint. It is better to detect a story that the team does not understand, or considers too large for a sprint, during refinement than during Sprint Planning.
- For the team, it is the chance to understand customer needs in advance, reduce surprises in Sprint Planning, and maintain a view of the product beyond the current iteration.
AI-assisted refinement
Generative AI tools can be used to support some phases of refinement:
- Story writing: based on meeting notes or user feedback, a language model can generate draft user stories that the Product Owner reviews and adjusts.
- Ambiguity detection: some teams use AI assistants to analyze stories before the session and identify acceptance criteria that might be missing or terms that could be interpreted in different ways.
- Dependency suggestions: in large backlogs, AI can help detect possible dependencies between stories that are not obvious at first glance.
These tools are useful as a starting point, but refinement is fundamentally a conversation between the team and the Product Owner. The shared understanding that this conversation generates cannot be replaced by an automatically generated document.
Common mistake
Treating refinement as an exhaustive requirements meeting. The goal of refinement is not to fully specify the stories, but to bring them to a level of detail sufficient for the team to estimate them and commit to them in the next sprint. Stories that are far off in time should have little detail; detail is added progressively as they get closer.
Resources
🏦 Skill User storiesState of the art and development guide · Skill Arena
📄 Product Owner v.2.0 [in Spanish]Free download · Scrum Manager
📄 Scrum Master v.4.0 [in Spanish] — Section: Product BacklogFree download · Scrum Manager
📄 Scrum Master en equipos con IA v.1.0 [Scrum Master in Teams with AI, in Spanish]Free download · Scrum Manager
See also
Want to advance in agility? You can search for course and exam calls or go at your own pace by becoming a member of the Agile Club. This membership includes exclusive resources, e-learning classrooms, and access to Skill Arena: a space to practice and measure your agile skills at your own pace.