Skip to content

RICE Assessment

RICE compares projects on the same basis. Three value dimensions divided by effort gives you an ROI-based grade.

What RICE does:

  • Apples-to-apples comparison across different project types
  • Makes value-effort tradeoffs visible
  • Gives prioritisation discussions a shared starting point

What RICE doesn’t do:

  • Automatically approve or reject projects
  • Determine when work gets scheduled
  • Override business judgement or strategic considerations

Applies to: initiatives scored in an RFC When to score: During Decide, include in the RFC Owner: Whoever owns the RFC


Score Reach, Impact, Confidence, and Effort (1–5 each). Divide.

RICE = (Reach + Impact + Confidence) / Effort

Round to one decimal place.

ScoreGradeWhat it means
5.0+AHigh ROI: strong value relative to effort
3.0–4.9BGood ROI: solid opportunity
1.5–2.9CModerate ROI: uncertain value or higher effort
<1.5DLow ROI: limited value or expensive to deliver

Grades describe ROI relative to other projects, not what to do with them. A C-grade project with high strategic importance may still be the right thing to build. Strategy decides what to work on. RICE sequences within a strategic bucket.


How broad is the audience?

ScoreDefinitionExample
1Niche: a handful of accounts with a specific needSpecific compliance requirement
2Segment: a defined customer segmentEnterprise-only feature
3Widespread: a large share of accounts, or a common secondary workflowHigh-usage reporting workflow
4Near-universal: most accounts or a core day-to-day workflowCore day-to-day workflow
5Universal: all accounts or all prospectsOnboarding, trial experience

💡 Ask yourself: is this a niche need, a segment need, a widespread need, a near-universal need, or a universal need?


How much does this change the outcome for each affected account?

This is depth, not breadth. Reach covers how many accounts are touched. Impact covers how much it matters to each one.

ScoreDefinitionSignalExample
1Nice-to-have: minor convenienceSmall friction reduction, no measurable metric movementClearer error formatting, UI polish
2Meaningful: measurably better workflowEfficiency gain, improved satisfactionFaster page loads, better in-context guidance
3Significant: removes a blocker, prevents churn, or unlocks a new capabilityBlocker removal, churn prevention, revenue impact per accountRemoves top deal-blocker, prevents at-risk churn
4Step-change: transforms the account’s relationship with the productNew revenue stream, 10x improvementA new capability that replaces a category of third-party tools; a foundational AI workflow
5Transformational: redefines the category or becomes core to the product’s identityMarket-creating, durable competitive moatA capability competitors can’t easily match; the reason accounts choose Tempo over anything else

💡 Would the customer say “that’s nice” (1), “that helps” (2), “that changes things for us” (3), “that changes everything” (4), or “that’s why we’re here” (5)?

Impact by strategic lever:

Lever12345
AcquisitionSlightly smoother trialFaster time-to-first-valueRemoves top deal-blockerOpens a new marketBecomes the category’s default reference product
RetentionReduces a minor annoyanceImproves satisfaction measurablyPrevents likely churnMakes switching unthinkableMakes the account structurally dependent on the product
ExpansionSmall usage growthDrives meaningful upsellUnlocks a new paid capabilityCreates a new revenue lineBecomes the primary driver of account growth
EfficiencySaves minutes per weekSaves hours per week or cuts costEliminates a significant burdenOrder-of-magnitude improvementRedefines how the whole org operates

How strong is the evidence?

ScoreDefinitionEvidence looks likeExample
1Hypothesis: team intuition, no external validationGut feel, analogies”We think customers want this”
2Directional: some signal, small sampleA few customer mentions, competitor has it3 support tickets, 1 customer interview
3Strong: multiple data points convergingMultiple customer requests, pipeline data, support patterns10+ requests, 5 deals blocked
4Validated: quantitative data with meaningful sampleUsage analytics, A/B results, survey with n>30Funnel data, NPS regression analysis
5Proven: causal evidence, replicatedControlled experiment, replicated across cohorts or time periodsA/B result confirmed in a second cohort, holds after a seasonality check

💡 Could you convince a skeptic with the evidence you have?

The Evidence & confidence sections in the Product Spec and Job Spec, and the Evidence section in the RFC, feed this score. More research and evidence there raises Confidence here; that is the loop from discovery to prioritisation.


How much work to deliver? Bigger effort = lower RICE score.

ScoreTeam commitmentExample
1~1 week, small changeUI tweak, config update
2~2 weeks, focused featureNew webhook event type
3~1 month, significant featureAnalytics MVP
4~3 months, major initiativeAuth system overhaul
5~6 months, platform-level changeNew compute backend

💡 How many sprints does this realistically tie up a squad?


ProjectRICValueEScoreGrade
Docs restructure42410110.0A
Trial limit changes3341033.3B
Analytics MVP333933.0B
Greenfield AI feature241741.8C
Niche integration112431.3D

If any dimension is unknown, mark as TBD. Grade cannot be calculated until all dimensions are scored.

RICE ranks projects within strategic buckets, not across them. Your annual strategy document sets this year’s strategic levers and capacity allocations.

Strategic levers:

  • Acquisition: Land and convert new customers
  • Retention: Keep customers, reduce churn
  • Expansion: Grow revenue from existing customers
  • Efficiency: Reduce costs, improve velocity

A B-grade retention project may beat an A-grade acquisition project if the retention bucket is under-invested. Strategy drives allocation. RICE drives sequencing within each bucket.

A project can score C on RICE and still have high absolute value, because RICE measures ROI (value per unit of effort). A 6-month platform project with universal reach and high impact still has moderate ROI per sprint. That doesn’t mean it’s wrong to build. Strategy and investment buckets handle those calls. RICE sequences the work within a bucket so you ship the highest-ROI items first.

If Confidence is 1 or 2, the grade carries more uncertainty:

  • Low confidence + high grade (A/B): Strong potential, but assumptions need testing
  • Low confidence + low grade (C/D): May not be worth the effort to validate further
  • High confidence + any grade: Scoring is well-grounded

Low confidence doesn’t automatically mean “don’t proceed.” It tells you the risk profile. Some strategic bets are worth taking despite uncertainty.


Every project must include a RICE Explanation that:

  1. Justifies Impact: Which strategic lever? What’s the magnitude of effect per account? What’s the evidence?
  2. Explains Confidence: What data supports this? What’s missing?
  3. Notes assumptions: What must be true for this score to hold?

Example (acquisition):

Impact 3 (significant): Removes a top-3 deal blocker cited by Sales. Accounts blocked by this can’t complete evaluation; it’s binary, they buy or they don’t. 15 active opportunities in pipeline waiting on this. Lever: Acquisition → Paid Conversion.

Confidence 3: Multiple customer requests + pipeline data. Would be 4 if we had quantitative trial drop-off data.

Example (retention):

Impact 2 (meaningful): Improves a frequently-used page’s performance at scale. Affected accounts save 5–10 minutes per user per day during triage. Two accounts are evaluating alternatives partly because of this. Lever: Retention.

Confidence 3: Customer interviews + support ticket volume. Would be 4 with usage analytics (see Product Analytics) on page load times.


PhaseRICE Status
LearnOptional: rough estimate helps early comparison
DecideRecommended: all dimensions scored for the RFC
BuildReference: update if scope changes significantly
LaunchRetrospective: compare predicted vs actual

Score before RFC approval when you can. The score is input to the decision, not a gate.

Re-score triggers:

  • Major scope change
  • New evidence that changes Confidence
  • Quarterly portfolio review