RICE Framework

RICE is a prioritization scoring model that ranks product opportunities using Reach × Impact × Confidence ÷ Effort, producing a numeric score across a backlog. It was developed at Intercom (Sean McBride, blogged 2017) and is used as one of the most common starting frameworks for product teams that want structured prioritization without inventing a custom model. It is widely adopted because it's simple enough to compute on a spreadsheet and rigorous enough to force the team to specify what they're claiming about each initiative.

The formula component by component: Reach is the estimated number of people (customers, users, requests) affected by the initiative in a defined time period (usually per quarter or per month), grounded in real numbers from analytics or CRM, not invented. Impact is the estimated effect on each affected person, scored on a small ordinal scale (Intercom's original: 3 = massive, 2 = high, 1 = medium, 0.5 = low, 0.25 = minimal), which deliberately resists false precision. Confidence is how sure the team is about the Reach and Impact estimates, also on a small scale (100% = high confidence, 80% = medium, 50% = low), used as a penalty for poorly-evidenced bets. Effort is the estimated person-months of work across the team. The formula puts everything together: a high-confidence initiative that reaches many people with strong impact and modest effort scores highest. Worked example: a feature reaching 5,000 customers per quarter with high impact (2) and high confidence (100%) at 2 person-months of effort scores (5000 × 2 × 1.0) ÷ 2 = 5,000. The 2024 to 2026 critique: RICE works best for incremental improvements to existing products where Reach and Effort can be estimated; it works worst for novel bets (where Reach is unknowable) and for strategic initiatives (where Impact isn't reducible to per-user impact). For those, use it alongside Cost of Delay, Opportunity Scoring, or strategic-bet frameworks.

Ryan's Take

RICE is useful exactly until founders start gaming it. The moment a team realizes the framework can be reverse-engineered to justify whatever the loudest stakeholder wanted, RICE stops being a decision tool and becomes a paperwork ritual. The defense against gaming is making the inputs public and the math auditable: if you scored Impact as 3 (massive) on a feature that affects 200 customers, somebody should be able to ask "show me which 200 customers and what kind of massive impact" and get a real answer. Frameworks only work in cultures where people will say "your Impact score is inflated."

What founders get wrong: Using RICE as a replacement for strategy rather than a tool inside one. A perfectly scored RICE backlog produces locally-optimal next features that may not ladder up to anything strategic. The framework ranks within a strategy; it doesn't replace the strategy. If you find RICE scores driving major direction shifts, the strategy upstream is doing too little of its job.

Related: [Feature Prioritization] · [Kano Model] · [Product Backlog] · [Product Management]

FAQ

What is the RICE framework? A prioritization scoring model developed at Intercom (Sean McBride, 2017) that ranks product opportunities using Reach × Impact × Confidence ÷ Effort. Produces a numeric score that surfaces relative priority across a large backlog. Widely adopted because it balances simplicity with rigor.

How do you calculate a RICE score? Multiply Reach (people affected per defined period) by Impact (small ordinal scale: 3 massive, 2 high, 1 medium, 0.5 low, 0.25 minimal) by Confidence (100% / 80% / 50%), then divide by Effort (person-months). Example: 5,000 customers × 2 (high impact) × 1.0 (high confidence) ÷ 2 person-months = 5,000.

When does RICE not work well? For novel bets where Reach is unknowable and for strategic initiatives where Impact isn't reducible to per-user impact. RICE works best for incremental improvements to existing products with measurable Reach. Pair with Cost of Delay or strategic-bet frameworks for the cases RICE handles poorly.