Requirements & constraints (clarify before scoring)
- Goal: prioritize work under ambiguous ROI + conflicting constraints (budget, time, regulatory).
- Outcome: single-page scorecard teams can apply consistently.
One-page framework (axes + brief rubric)
- Customer Impact (0–10)
- 0: no user benefit; 5: measurable retention/engagement uplift; 10: core user problem / revenue driver.
- Risk & Compliance (0–10, reversed for risk)
- 0: major regulatory/legal risk; 5: manageable with controls; 10: no compliance issues.
- Implementation Effort (0–10, reversed for cost/time)
- 0: >3 months / large cross-team work; 5: 3–6 weeks; 10: <1 week / low effort.
- Strategic Alignment (0–10)
- 0: misaligned; 5: adjacent; 10: central to company OKRs/roadmap.
- Financial/ROI Confidence (0–10)
- 0: no ROI data; 5: proxy metrics; 10: validated projection with sensitivity analysis.
Scoring approach
- Normalize: higher is better for all axes (reverse-score Risk and Effort).
- Weighted sum: Final Score = 0.30Customer + 0.20Risk + 0.20Effort + 0.15Strategic + 0.15*ROI.
- Thresholds: >8 — green (build), 5–8 — yellow (pilot/limited), <5 — red (defer).
- Tie-breaker: prefer lower implementation effort if scores within 0.2.
How to score consistently
- Short rubric examples per axis on the one-pager (2–3 bullet anchors).
- Require evidence field for each score: metric, assumption, or doc link (e.g., user research, cost estimate, legal memo).
- Owner signs off: Product proposes scores; Tech lead verifies effort; Compliance verifies risk.
Transparency & governance
- Shared single source: store scorecards in central board (Jira/Notion/Confluence template).
- Weekly prioritization sync: review new/changed items with product, eng lead, and compliance.
- Audit trail: track score changes, who changed them, and rationale.
- Re-evaluation cadence: run for long-lived items every sprint or after new data.
- Calibration sessions: monthly 30-min cross-team scoring calibration using 2 sample items to align interpretation.
Practical notes for engineers
- Keep effort estimates small (T-shirt or week buckets) to reduce false precision.
- When ROI is ambiguous, capture assumptions and sensitivity (best/worst case).
- Use this as a decision aid, not a bureaucratic gate—fast-path smaller items with high score + low effort.
This one-page balances quantitative scoring with required qualitative evidence to keep decisions fast, auditable, and aligned across product and engineering.
