Epic Breakdown
Break a large product initiative or PRD into well-defined epics. Each epic is independently shippable, estimable in story points, and clearly tied to a business outcome.
When to use
- Just finished a PRD and need to plan implementation
- A roadmap initiative is too big to start estimating
- Sprint planning keeps stalling because work isn't broken down enough
- Need a release plan to share with stakeholders
Prompt
You are a senior product manager and agile coach. Break the initiative below
into well-scoped epics suitable for multi-sprint delivery.
## Input
**Initiative or PRD:**
{{initiative}}
**Target planning horizon:** {{target_quarters}} quarter(s)
## Epic definition
Each epic must be:
- **Outcome-oriented** — tied to a measurable business or user outcome
- **Independent** — deliverable without requiring other epics to ship first
(or with explicit dependencies called out)
- **Estimable** — sized in t-shirt: S (1-2 sprints), M (3-4 sprints), L (5+ sprints)
- **Releasable** — produces something users actually experience, not just internal plumbing
- **Vertical** — slices through all layers (UI, API, DB) rather than horizontal layer-only
## What to produce
### 1. Initiative summary
- 2-3 sentence summary of what's being built and why
- Success metrics (how we'll know it worked)
- Key constraints (timeline, team size, tech)
### 2. Epic list
For each epic produce:
Epic: [Outcome-oriented title]
ID: EPIC-001 Size: S | M | L Priority: P0 | P1 | P2 Outcome: [Measurable outcome this epic delivers] User benefit: [What changes for the user] Business value: [What changes for the business]
Scope
What's IN:
- [Bulleted list of capabilities]
What's OUT:
- [Explicit non-goals — things stakeholders might assume are in but aren't]
Stories (estimate-only, full breakdown later)
- ~ STORY: [one-liner], complexity: S/M/L
- ~ STORY: [one-liner], complexity: S/M/L
- (5-12 stories per epic typical)
Dependencies
- Blocks: [other epics that need this first]
- Blocked by: [epics this needs first]
- External: [other teams, vendors, or systems]
Risks
- [Top 1-3 risks: technical unknowns, regulatory, vendor, etc.]
Definition of done
- [3-5 measurable acceptance criteria for the epic as a whole]
### 3. Sequencing recommendation
- Suggested order with rationale
- Group epics into proposed sprints/milestones
- Note critical path and what determines it
### 4. Out of scope
Things that came up while breaking down but explicitly aren't part of THIS
initiative. (Future-roadmap candidates.)
### 5. Open questions
Decisions still needed before work can start. For each:
- The question
- Who needs to answer
- Default assumption if no one answers in time
- Cost of the wrong default
## Quality bar
- Every epic ties to a user or business outcome (no "Refactor module X" epics)
- Every epic could ship independently (or has a clear dependency note)
- No epic is bigger than 6 sprints — if it is, split it
- No epic is smaller than 1 sprint — if it is, it's probably a story, fold into another epic
- Vertical slices, not horizontal layers ("Add login flow" not "Set up auth tables")
## Output format
Markdown document with:
- Title
- Initiative summary
- Visual sequence diagram (ASCII art or Mermaid)
- Each epic as a section
- Sequencing recommendation
- Out of scope
- Open questions
## Tone
- Outcome-focused
- Specific over generic
- Honest about unknowns
- Numbers, not adjectivesExample output snippet
## Epic: Self-service invoice management
**ID:** EPIC-001
**Size:** M (3 sprints)
**Priority:** P0
**Outcome:** Reduce account-manager invoice requests by 60%
**User benefit:** Customers download invoices in under 30 seconds, 24/7
**Business value:** ~80 hours/week of account-manager time freed
### Scope
What's IN:
- View invoices for last 12 months
- Download PDF
- Filter by date range
- Email-yourself-a-copy action
What's OUT:
- Invoice editing or disputes (separate epic)
- Bulk export beyond 12 months (separate)
- Tax document generation (separate epic)Tips
- Run this AFTER the PRD is solid — breakdown quality depends on PRD quality
- Pair with the User Story Generator to drill into the highest-priority epic first
- Pair with the Jira Ticket Manager Skill to bulk-create epics in your tracker
- Revisit the breakdown every 2-3 sprints — reality will diverge from the plan
Common mistakes to avoid
- "Plumbing" epics with no user-facing outcome (fold infrastructure work into the user-facing epics it enables)
- Epics that are really just oversized stories (size: S epics are probably stories)
- Epics with vague success criteria
- Skipping the "out of scope" section — it prevents scope creep