RICE

What it is

RICE is a prioritization framework created by Sean McBride on Intercom's growth team (2018) to settle recurring arguments about what to build next by replacing gut feel with a consistent, comparable score. It rates each initiative on four factors — Reach, Impact, Confidence, Effort — and combines them as (Reach × Impact × Confidence) ÷ Effort, yielding a single number that approximates "total impact per unit of work." It matters because it forces you to separate a project's upside (Reach × Impact) from your certainty about that upside (Confidence) and its cost (Effort) — three things teams routinely conflate — so a well-defined small win can outrank an exciting but speculative moonshot.

When to use it

• Use it in prioritization and roadmap questions ("You have 10 features and one quarter — how do you decide?"), when you must compare heterogeneous initiatives on a level playing field, or when a stakeholder is championing a pet project and you need an objective, defensible ranking.
• It shines when you have (or can estimate) real usage metrics for Reach.

When not to use: Skip it for a small set of strategically obvious must-dos, for genuine 0-to-1 bets where Reach and Impact are pure guesses (the math just launders low confidence into false precision), or when items are tightly coupled by dependencies that a single score can't capture.

The steps

1. 1
   Define the scope and time windowList the candidate initiatives and fix one consistent time period for Reach (e.g. 'users per quarter') so every project is measured on the same clock — mixing per-month and per-quarter reach silently distorts the ranking.
2. 2
   Estimate Reach with real numbersCount how many people or events the initiative touches per period using actual product metrics, not vibes — e.g. 'this checkout change is seen by 40,000 of our 120,000 monthly buyers' — and score the affected segment, not your total user base.
3. 3
   Assign Impact from the fixed scaleRate how much it moves the goal per person using Intercom's five discrete tiers — Massive=3, High=2, Medium=1, Low=0.5, Minimal=0.25 — and resist inventing values like 1.7; the coarseness is deliberate to prevent false precision.
4. 4
   Apply Confidence as a discount for uncertaintyMultiply by how much you trust your own estimates using Intercom's three tiers — High=100%, Medium=80%, Low=50% — where a Low score flags an idea you should probably de-risk before it earns roadmap space; this is the lever that stops a data-free bet from topping the list.
5. 5
   Estimate Effort in person-monthsSum the total product + design + engineering work in person-months (count anything under a month as 0.5), then remember that because Effort is the denominator, it's the one factor where a bigger number lowers the score.
6. 6
   Compute, rank, then sanity-checkCalculate (R × I × C) ÷ E for each, sort descending, and pull the ranking up to gut-check it against dependencies and strategy — RICE surfaces trade-offs but shouldn't override judgment, so be ready to explain any manual reorder.

Worked examples

Common mistakes

• Scoring Reach against your total user base instead of the segment the feature actually touches, which inflates broad-but-shallow features over targeted high-impact ones.
• Chasing false precision — debating whether Impact is 1.3 or 1.7 — when the framework is deliberately coarse and only offers 3 / 2 / 1 / 0.5 / 0.25.
• Chronically underestimating Effort (forgetting design, QA, and cross-team coordination), which is especially damaging because Effort is the denominator and small errors swing the ranking.
• Treating the number as the decision rather than an input — shipping the top score even when a dependency, strategic bet, or compliance deadline should override it.