A four-factor prioritisation formula developed at Intercom that scores candidates by (Reach × Impact × Confidence) ÷ Effort to produce a defensible, comparable ranking.
Which candidate will deliver the most value to the most users, given how confident we are and how much it costs to build?
| Item | Reach | Impact | Confidence | Effort | Score▼ |
|---|---|---|---|---|---|
Template Libraryfeature | 5 | 3 | 4 | 1 | 60 |
Export to PDFfeature | 5 | 2 | 5 | 1 | 50 |
AI Copilotfeature | 4 | 5 | 3 | 2 | 30 |
Custom Dashboardsfeature | 2 | 4 | 3 | 4 | 6 |
Real-time Collabfeature | 3 | 4 | 2 | 5 | 4.8 |
(reach * impact * confidence) / effortRICE is a prioritisation formula. It scores each candidate item by four factors, Reach, Impact, Confidence, and Effort, and produces a single number that ranks the candidate against others. The ranking is the starting point for a conversation, not the conclusion. RICE makes the assumptionsAssumptionStrategyA belief taken as true that underpins a strategyView reference → behind a prioritisation decisionDecisionStrategyA recorded decision with context, rationale, and consequencesView reference → explicit and comparable, which is its main value.
Sean McBride at Intercom developed RICE after the team had been working with the simpler ICE framework (Impact, Confidence, Effort) and found it systematically underweighted audience size. A featureFeatureProduct SpecificationA product capability or featureView reference → affecting five percent of users and a feature affecting the entire user base could score identically under ICE if the other two factors matched. Adding Reach as a multiplier corrects that. McBride published the framework on the Intercom blog around 2016 in a post titled "RICE: Simple prioritization for product managers," which remains the canonical reference. The framework was adopted quickly across SaaS product teams and became a native scoring method in tools including Productboard, Aha!, and various Notion templates. The formula and its four components have not changed materially since the original post; most teams vary their scoring scales for Impact and Confidence to match their own evidenceEvidenceValidationData supporting or refuting a hypothesisView reference → norms.
The formula is: RICE score = (Reach × Impact × Confidence) ÷ Effort.
Reach is the number of users or customers the feature will affect in a given period, typically one quarter. Use counts of real users, not pageviews or impressions.
Impact is a score representing how significantly the feature moves the needle for each affected user. Intercom's original scale uses 3 (massive), 2 (significant), 1 (moderate), 0.5 (low), and 0.25 (minimal). Some teams simplify to 1, 2, or 3.
Confidence is a percentage expressing how certain the team is about the Reach and Impact estimates. 100% means there is strong evidence (user research, quantitative data). 80% means there is partial evidence. 50% means it is mostly a hypothesisHypothesisValidationA testable belief about a solutionView reference →.
Effort is the total person-weeks of work required across all disciplines, not just engineering.
A worked example. The team is choosing between two candidates.
Candidate A, an in-app notification system: Reach = 800 users per quarter, Impact = 2 (significant), Confidence = 80%, Effort = 4 person-weeks. RICE = (800 × 2 × 0.8) ÷ 4 = 1,280 ÷ 4 = 320.
Candidate B, a bulk-export feature: Reach = 200 users per quarter, Impact = 3 (massive), Confidence = 50%, Effort = 2 person-weeks. RICE = (200 × 3 × 0.5) ÷ 2 = 300 ÷ 2 = 150.
Candidate A ranks higher. The bulk-export feature has higher per-user impact and lower effort, but its narrow audience and low confidence pull its score below the notification system. The team now has a number to argue with, and a record of the assumptions behind it.
RICE works well when the candidate list is long enough that triage is the real problem, and when candidates genuinely differ in how many users they affect. B2C product teams and growth teams commonly reach this condition: dozens of plausible items, a meaningful spread in audience size, and a needNeedUserA user need, pain, desire, or constraintView reference → to defend prioritisation decisions to stakeholdersStakeholderTeam & OrganisationA person with influence over the productView reference →.
It is less suited to two situations. In B2B enterprise product development, nearly every feature affects the same small set of enterprise accounts, so Reach loses its discriminating power and the formula collapses toward ICE. In early-stage products with no usage data, Reach estimates are guesses, Confidence is structural 50%, and the output precision flatters what is actually a shallow ranking.
The failure modes worth knowing: treating Confidence as 100% by default because it is optimistic, which inflates every score equally and wastes the factor's value; estimating Effort in calendar time and not person-weeks, which produces inconsistent denominators; and arguing about RICE scores to two decimal places, which mistakes precision in arithmetic for precision in judgement. Round to whole numbers. The ranking is what matters, not the absolute score.
RICE is a table framework in the prioritisation category. Its rows are items under consideration, and its columns are the four scoring properties. In the Unified Product Graph, each row maps to a feature
featurefeaturefeaturefeaturefeatureStoring the four scores as properties on the feature