Feature Prioritization

Feature Prioritization

Purpose

Prioritize a feature backlog using a structured scoring framework (RICE by default) and produce a ranked list with rationale.

Input: Feature list, product vision or goals (optional), constraints (timeline, team size) Output: Prioritized feature ranking with scores, trade-off notes, and recommended MVP cut line Examples: See [examples.md](examples.md) for worked input/output.

---

Workflow

Step 1: Normalize feature list

For each feature, extract:

• Name and one-sentence description
• User/problem it addresses
• Dependencies on other features
• Estimated effort category (S/M/L/XL if no estimates provided)

If features are vague, ask for clarification before scoring.

Step 2: Select scoring framework

Default: RICE (Reach, Impact, Confidence, Effort)

| Framework | Use when | |-----------|----------| | RICE | Quantitative prioritization with multiple features | | ICE | Quick prioritization, early stage, less data | | MoSCoW | Release planning with fixed deadline |

State which framework is used and why.

Step 3: Score each feature

For RICE:

| Factor | Scale | Guidance | |--------|-------|----------| | Reach | # users per quarter | Estimate from persona/segment | | Impact | 0.25–3 | Massive=3, High=2, Medium=1, Low=0.5, Minimal=0.25 | | Confidence | 0–100% | Evidence quality | | Effort | person-months | Team effort estimate |

RICE Score = (Reach × Impact × Confidence) / Effort

Step 4: Rank and draw MVP line

• Sort by RICE score descending
• Draw MVP cut line based on stated constraints
• Document features below the line and why deferred

Step 5: Document trade-offs

For top 5 features, note:

• What is sacrificed by prioritizing this
• Dependencies and sequencing requirements
• Risks of deprioritizing alternatives

Step 6: Validate

Run Validation checklist.

---

Decision Rules

| Condition | Action | |-----------|--------| | Feature list empty | Stop; request features or reference PRD | | No reach data available | Use relative reach estimates; mark Confidence lower | | Effort unknown for all features | Use t-shirt sizes converted to person-months; flag estimates | | Two features are mutually exclusive | Score separately; add decision note to pick one | | Constraint is fixed date | Use MoSCoW instead of RICE; mark Must/Should/Could/Won't |

---

Validation

• [ ] Scoring framework explicitly stated
• [ ] Every feature has all scoring dimensions filled
• [ ] RICE/ICE scores calculated correctly
• [ ] Ranked list matches calculated scores
• [ ] MVP cut line drawn with rationale
• [ ] Top 5 trade-offs documented
• [ ] Dependencies noted between features
• [ ] No features scored without rationale

---

Anti-patterns

• HiPPO prioritization — ranking by stakeholder loudest voice without scores.
• Everything is Must — MoSCoW with no Could/Won't items.
• Fake precision — confidence 100% without evidence.
• Ignoring dependencies — ranking independent of sequencing constraints.
• Effort-free ranking — high impact features that take a year ranked first.

---

Best Practices

• Align Impact scores with north star metric from product vision.
• Lower Confidence when based on assumptions; note what would raise it.
• Include tech debt and infrastructure items in the same framework.
• Re-score after major new evidence (user research, prototype results).
• Document what was explicitly deprioritized and why.

---

Output Structure

# Feature Prioritization: [Product Name]

## Framework
**Method:** RICE | ICE | MoSCoW
**Rationale:** [why this framework]

## Constraints
- Timeline: [if stated]
- Team: [if stated]

## Scored Features
| Rank | Feature | Reach | Impact | Confidence | Effort | Score | Rationale |
|------|---------|-------|--------|------------|--------|-------|-----------|

## MVP Cut Line
**Included in MVP:** features #1–#N
**Rationale:** [why this cut]

## Deferred Features
| Feature | Reason deferred |
|---------|-----------------|

## Trade-offs (Top 5)
| Feature | Gain | Sacrifice |
|---------|------|-----------|

## Dependencies
[Feature A] → requires → [Feature B]

---

Next Skills

| Outcome | Recommended Skill | |---------|-------------------| | MVP scope defined | product/prd-generator | | Need epic breakdown | product/epic-generator | | Need user stories | product/user-story-generator | | Vision unclear | product/product-vision-builder |

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: patonkikh
• Source: patonkikh/APES
• License: MIT

Install and usage instructions live in the source repository linked above.