Purpose
Turn a messy list of competing initiatives into a defensible, weighted priority ranking — with transparent scores, tiered groupings, and a clear sequencing recommendation — so the team stops arguing about opinions and starts executing against a shared model.
When to Use
Invoke this skill when the user asks to:
- Prioritize a product backlog or list of feature ideas
- Decide what to build next sprint, quarter, or half
- Rank initiatives when there are more requests than capacity
- Justify a roadmap decision to leadership or stakeholders
- Resolve disagreement about what should come first
- Re-sequence a roadmap after a strategy shift or resourcing change
Inputs Required
Before starting, collect:
- Initiative list: The candidate initiatives, features, or projects to rank (name + 1-sentence description each)
- Strategic goals: The 2–4 company or product objectives this roadmap serves this period
- Time horizon: The window being planned (e.g., next quarter, next 6 months)
- Team capacity: Rough available effort — team size, sprint count, or engineering-weeks
- Criteria weights: How to weight impact, effort, strategic fit, and urgency (or accept defaults)
- Known constraints: Hard dependencies, committed deadlines, or non-negotiables
- Effort estimates: If available — rough sizing per initiative (T-shirt or story points)
If weights are not provided, default to Impact 35%, Strategic Fit 25%, Urgency 20%, Effort 20% (effort scored inversely — lower effort scores higher). If effort estimates are missing, estimate them and flag the assumption.
Instructions
Step 1: Normalize the Initiative List
Restate each initiative as a clear, comparable unit of work:
- Give each a short label and a one-line outcome statement ("what changes if we ship this")
- Split any initiative that bundles two distinct outcomes into separate rows
- Flag anything too vague to score and note what information is missing
- Number the initiatives for reference throughout
Step 2: Confirm the Scoring Model
Establish the four criteria and their weights before scoring anything:
- Impact (0–5): Expected effect on the strategic goals — revenue, retention, activation, cost, or risk reduction. Score against the goals provided, not generic value.
- Effort (0–5, inverse): Estimated cost to deliver. Score 5 = trivial, 0 = massive multi-quarter build. This dimension rewards leverage.
- Strategic Fit (0–5): How directly the initiative advances the stated strategic goals versus being a side quest.
- Urgency (0–5): Time sensitivity — deadlines, competitive pressure, decaying opportunity, or compounding cost of delay. Score 5 only when delay carries real cost.
State the active weights explicitly (defaults or provided) so the math is auditable.
Step 3: Score Each Initiative
For every initiative, assign a 0–5 score on all four criteria with a one-clause justification per score (not just a number). Then compute:
Weighted Score = (Impact × w₁) + (Effort × w₂) + (Strategic Fit × w₃) + (Urgency × w₄)
Report each initiative's total on a normalized 0–100 scale for easy comparison. Be consistent — if two initiatives have similar impact, their scores should reflect that.
Step 4: Rank and Tier
Sort by weighted score descending, then group into tiers:
- Tier 1 — Do Now: Highest scores; clear leverage and fit. Should fit within stated capacity.
- Tier 2 — Do Next: Strong but capacity- or dependency-constrained this horizon.
- Tier 3 — Deprioritize / Revisit: Low leverage, weak fit, or premature — with a reason each was cut.
- Quick Wins callout: Flag any high-impact, low-effort items regardless of tier — these often jump the queue.
Step 5: Apply Constraints and Sequence
Overlay the real-world constraints on the pure scores:
- Reorder for hard dependencies (prerequisite work moves earlier even if scored lower)
- Honor committed deadlines and non-negotiables
- Check the Tier 1 set against team capacity — cut the lowest-scoring item that overflows and note it
- Note any two initiatives that should ship together or are mutually exclusive
Step 6: Deliver the Recommendation
State a clear sequencing recommendation for the horizon: what to build, in what order, and what got deferred. Call out the single highest-leverage bet and the biggest thing being consciously cut. Be opinionated — the team needs a decision, not a menu.
Output Format
- Structure: Scoring model summary → ranked scoring table → tier groupings → sequencing recommendation
- Length: 700–1,100 words plus the scoring table
- Format: Markdown table for scores (columns: #, Initiative, Impact, Effort, Fit, Urgency, Weighted Score, Tier), prose for the recommendation
- Sections: Scoring Model & Weights → Ranked Scoring Table → Tier Groupings (with quick-wins callout) → Sequencing Recommendation
- Tone: Product leader — decisive, transparent about trade-offs, willing to name what got cut and why
Quality Checklist
Before delivering output, verify:
- [ ] Every initiative from the input appears in the scoring table (count them)
- [ ] All four criteria are scored 0–5 for every initiative with a justification per score
- [ ] The active weights are stated explicitly and the weighted math is consistent with them
- [ ] Effort is scored inversely (low effort = high score) — not confused with impact
- [ ] Tier 1 respects stated team capacity, or overflow is explicitly flagged
- [ ] The final recommendation names both the top bet and what was consciously deprioritized
Notes
- Effort scores are the most common source of error — anchor them to a shared unit (weeks or points) before scoring to keep them comparable
- Quick wins (high impact, low effort) frequently outrank higher-scored initiatives in practice; ship them early to build momentum
- When two initiatives tie, break the tie with strategic fit — it ages better than urgency
- Re-run this whenever strategy shifts; a stale ranking is worse than none because it carries false confidence
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills