10.3 Refine Prioritized Product Backlog
| 10.3 Refine Prioritized Product Backlog | ||
|---|---|---|
| Inputs | Tools | Outputs |
Bold ITTOs are mandatory.
Refine Prioritized Product Backlog is the ongoing process of clarifying, splitting, estimating, and reordering backlog items so the highest-value work is ready for upcoming sprints.
Purpose & When to Use
The goal is to keep the Prioritized Product Backlog healthy—clear, estimated, and ordered—so Sprint Planning is fast and predictable. Refinement ensures that items near the top are small, testable, and understood by the Scrum Team.
- Use continuously throughout the project; commonly 5–10% of team capacity each sprint.
- Schedule regular sessions mid-sprint; trigger ad-hoc refinement when new information, risks, or regulatory changes appear.
- Product Owner leads content and ordering; Scrum Master facilitates; Developers/Team provide technical insight and estimates; stakeholders join as needed for clarity.
Mini Flow (How It’s Done)
- Prepare inputs: current Prioritized Product Backlog, feedback from reviews, latest metrics, risks, and constraints.
- Select top candidates: choose items likely to enter the next 1–2 sprints.
- Clarify value and users: restate the problem, business value, and user impact.
- Add or improve acceptance criteria: define objective, testable outcomes.
- Split large items (epics) into thin, independent slices that can finish within one sprint.
- Discuss dependencies, assumptions, and non-functional needs (performance, security, compliance).
- Estimate collaboratively (e.g., story points or T-shirt sizes) using relative sizing.
- Reorder based on value, risk, cost, and learning opportunity; remove or archive obsolete items.
- Check readiness: confirm items meet the team’s Definition of Ready for Sprint Planning.
- Update artifacts: risks, release forecasts, and any architectural spikes or follow-ups.
Quality & Acceptance Checklist
- Each near-term item has a clear title, short description of value, and target user.
- Acceptance criteria are specific, testable, and include key non-functional needs.
- Item is small enough to complete within one sprint without hidden work.
- Team-provided estimate exists and is credible relative to similar work.
- Dependencies, constraints, and assumptions are known and visible.
- Priority reflects current value, risk, and time sensitivity; obsolete items removed.
- Relevant risks captured and linked to the item; mitigation notes added if needed.
- Test and demo approach is understood; traceability to goals or releases is clear.
- Item meets the team’s Definition of Ready; if not, specific next steps are recorded.
Common Mistakes & Exam Traps
- Treating refinement as optional; outcome is chaotic Sprint Planning and carryover.
- Over-refining low-priority items while near-term items remain vague.
- Product Owner doing it alone; estimates must come from the team.
- Turning refinement into a technical design meeting; keep it about clarity and size.
- Confusing refinement with Sprint Planning; no commitment is made during refinement.
- Skipping acceptance criteria or non-functional needs; leads to rework and defects.
- Changing in-sprint scope without following change control or renegotiation.
- Not splitting epics; oversized items cannot be finished in a single sprint.
PMP/SCRUM Example Question
During a mid-sprint refinement session, the team discovers that the top backlog item is too large and lacks acceptance criteria. What should the Scrum Team do first?
- Estimate it quickly so it can be pulled into the next sprint.
- Ask the Product Owner to split the item and add acceptance criteria, then re-estimate.
- Remove the item from the backlog until all details are known.
- Assign a developer to create a detailed technical design before splitting.
Correct answer: B. The proper refinement action is to collaborate with the Product Owner to split the item into smaller, valuable slices and add clear acceptance criteria, then re-estimate and reorder as needed; estimation before clarity (A) is unreliable, removing the item entirely (C) loses value, and doing detailed design (D) is beyond refinement’s scope.
HKSM