Compass← All frameworks

Framework

RICE Scoring

Reach × Impact × Confidence ÷ Effort.

Best for
Prioritising a product or work backlog
Time needed
30 minutes
Origin
Intercom's product team, ~2016.

RICE turns 'what should we build?' into a comparable score. It's not precise — the point is a shared, transparent ranking that surfaces disagreement about assumptions rather than opinions.

Use it when

  • A team is arguing about roadmap priorities.
  • You have more than five candidates.
  • You want to make the tradeoffs visible.

Skip it when

  • Strategic bets that don't fit the model (a moonshot has low confidence by definition).
  • One-off personal decisions — use pros/cons instead.

The steps

  1. 01

    Reach

    How many people affected per time period? (Use a real number.)

  2. 02

    Impact

    Massive 3, high 2, medium 1, low 0.5, minimal 0.25.

  3. 03

    Confidence

    100% high evidence, 80% medium, 50% low. Below 50%, reconsider.

  4. 04

    Effort

    Person-months of work.

  5. 05

    Compute and rank

    Score = (Reach × Impact × Confidence) ÷ Effort. Sort descending.

A worked example

"Which of five features should we build next quarter?"

Feature A: 5000 users × 2 impact × 80% conf ÷ 2 months = 4000. Feature B: 500 × 3 × 100% ÷ 1 = 1500. A wins.

Outcome

Build A. Revisit B when we have more confidence data.

Where it quietly breaks

  • Inflating confidence to reach a desired ranking.
  • Treating the score as truth instead of a shared conversation.

Run this on a real decision — in Compass.

Begin free

Related frameworks

Eisenhower Matrix

Sort every task by urgency and importance.

Cost-Benefit Analysis

Put a number on every side of the decision.