RICE: end the loudest-voice roadmap
Intercom's scoring framework as its authors defined it — the four factors, the fixed scales that keep scores honest, and the confidence discount that punishes wishful thinking.
The theory in one paragraph
Every roadmap meeting has a gravitational pull toward whoever argues best — the founder's pet feature, the loudest customer, the deal sales swears will close. RICE, developed inside Intercom's product team, replaces advocacy with four estimates: Reach (how many customers this touches per quarter), Impact (how much it moves the goal for each of them), Confidence (how much evidence backs those numbers), and Effort (person-months to ship). Score = Reach × Impact × Confidence ÷ Effort. The formula isn't the point — the point is that every input is a claim someone can challenge with data, and the confidence multiplier forces you to admit which numbers you made up.
How it works
The mechanics — as Sean McBride defined them, not the folklore version.
Reach and Impact — size times depth
Reach is a real count over a real period: customers per quarter, signups per month — pulled from analytics, not vibes. Impact is how strongly the change moves your goal for each person reached, on a fixed scale: 3 massive, 2 high, 1 medium, 0.5 low, 0.25 minimal. The fixed scale is deliberate — free-form impact numbers turn the exercise into whoever exaggerates best.
Confidence — the honesty discount
100% means you have data for both reach and impact; 80% means data for one and informed judgment on the other; 50% means it's mostly hypothesis. Anything below 50% is what McBride called a moonshot — score it, but know you're gambling. Confidence is the factor that separates RICE from its predecessors: it makes uncertainty a visible, discounted input instead of a hidden assumption.
Effort — the denominator that kills epics
Person-months across product, design, and engineering, estimated in whole numbers (0.5 for true quick wins). Because effort divides rather than subtracts, doubling the size of a project halves its score — which is exactly the pressure that pushes teams to find the two-week version of the six-month idea.
The score is a conversation starter, not a verdict
Teams that treat RICE output as an auto-sorted roadmap miss where the value lives: in the arguments the inputs force. When two people score the same feature 3× apart, the disagreement is always about a hidden assumption — who it's for, what goal it serves, what's actually being built. Intercom's own guidance is to re-order the list after scoring when strategy demands it; the framework's job is to make you say out loud why you're overriding the math.
The person behind it
Sean McBride & the Intercom product team
Product managers at Intercom · published the framework on Inside Intercom, 2016
McBride co-developed RICE while Intercom's product team was outgrowing gut-feel prioritization and existing scoring models kept rewarding whoever estimated most optimistically. The fix that made RICE spread — fixed multiple-choice scales for impact and confidence instead of free-form numbers — is precisely what keeps it from degenerating into arithmetic-flavored opinion.
How to apply it this week
Each step maps to a field in the Development Roadmap tool — finishing the read means finishing the work.
Fix the goal before you score anything
RICE scores are relative to one metric — activation, retention, revenue. A backlog scored against 'growth' in general lets every feature claim impact on something. Write the quarter's goal at the top of the sheet.
Development Roadmap · goalsList candidates at comparable altitude
Score features against features, projects against projects. Mixing 'redesign onboarding' with 'fix tooltip copy' produces scores that flatter the tiny and punish the ambitious for their honesty about effort.
Development Roadmap · features & milestonesEstimate reach from your own analytics
Count the actual users who hit this flow, segment, or pain per quarter. If you can't pull the number, that's not a scoring problem — it's a signal your confidence input just dropped to 50%.
Score impact and confidence on the fixed scales
Impact: 3 / 2 / 1 / 0.5 / 0.25. Confidence: 100 / 80 / 50%. Resist inventing intermediate values — the coarseness is what keeps three people's scores comparable and the meeting short.
Rank, argue the outliers, then commit the order
Sort by score, then interrogate surprises: a beloved project scoring low usually has invented reach; a boring one scoring high is usually a real quick win. Document any override and its reason — that log is your prioritization strategy, written one honest exception at a time.
Development Roadmap · timeline
Build it, don't just read it
The steps above are the Development Roadmap tool's structure. Open it and work through them with your own startup — your readiness score starts building from the first field.
Free account · no card required
See it in the wild
Teardowns from our benchmarks library where this framework is doing real work.
Benchmark teardown
Careem
Super-app expansion is a RICE exam: every candidate vertical scored on reach of the existing rider base versus real build effort
Read the teardown
Benchmark teardown
Duolingo
An experiment culture where reach and impact come from A/B data — confidence at 100% is the house style
Read the teardown
Benchmark teardown
Spotify
The podcast bet reads as a low-confidence, massive-impact moonshot — scored and taken deliberately, not by formula
Read the teardown
Frequently asked questions
What is the RICE prioritization framework?
RICE is a scoring model from Intercom's product team for ranking competing ideas: Reach (how many customers it affects per time period) × Impact (how much it moves your goal per customer, on a 0.25–3 scale) × Confidence (how much evidence supports your estimates: 100%, 80%, or 50%) ÷ Effort (person-months to build). Higher scores suggest more value per unit of work.
How do you calculate a RICE score with an example?
Say a feature reaches 500 customers a quarter, with high impact (2), backed by data for reach but judgment on impact (80% confidence), needing two person-months: 500 × 2 × 0.8 ÷ 2 = 400. A rival idea reaching 5,000 customers at minimal impact (0.25), full confidence, one month scores 5,000 × 0.25 × 1 ÷ 1 = 1,250 — the unglamorous broad fix beats the deep niche feature.
What is the difference between RICE and ICE scoring?
ICE (Impact, Confidence, Ease — popularized by Sean Ellis for growth experiments) is faster: three gut-scored factors, built for weekly experiment queues. RICE adds Reach as an explicit, data-derived factor and replaces Ease with Effort as a divisor, which makes it better suited to roadmap decisions where audience size and engineering cost differ wildly between candidates.
What are the weaknesses of RICE prioritization?
Garbage in, garbage out: invented reach numbers and optimistic impact scores produce confident-looking nonsense. It also structurally favors incremental work — big strategic bets score low on confidence and high on effort — so most teams pair it with a deliberate moonshot allocation. And it can't see dependencies or strategy; the score ranks value density, not sequencing.
When should a startup use RICE instead of just deciding?
When the backlog outgrows what one person holds in their head, when prioritization debates keep being won by seniority or volume, or when saying no to stakeholders needs a shared, legible rationale. Pre-product-market-fit, a lightweight ICE pass is usually enough; RICE earns its overhead once real usage data exists to feed the reach input.
Related frameworks
OKRs
Andy Grove & John Doerr · 1999
North Star Metric
Sean Ellis · 2017
The Lean Startup
Eric Ries · 2011
Sources
Independent educational summary written by StartupKit from public sources. RICE Prioritization is the work of Sean McBride & the Intercom product team; this page is not affiliated with or endorsed by the author.