How to Create a RICE Scoring Spreadsheet for Product Prioritization
RICE scoring — Reach, Impact, Confidence, Effort — is a framework developed by Intercom PMs to bring structure to backlog prioritization. It reduces the "loudest voice wins" dynamic by forcing every feature to be evaluated on the same four dimensions, producing a single numeric score you can sort by.
This guide walks through building a RICE scoring spreadsheet from scratch, including formulas, calibration techniques, and how to use the output to make defensible prioritization decisions.
Step-by-step guide
Set up the spreadsheet structure
Create a Google Sheet with these columns: Feature Name (text), Reach (number: users affected per quarter), Impact (scale: 3=massive, 2=high, 1=medium, 0.5=low, 0.25=minimal), Confidence (percentage: 100%=sure, 80%=high, 50%=medium, 20%=low), Effort (person-months), RICE Score (formula), and Notes. Freeze the header row and the Feature Name column (View > Freeze > 1 row, 1 column).
Define your Reach calibration
Reach measures how many users the feature will affect in a specific time period — use one quarter consistently. Define reach in terms of your actual user base. If you have 10,000 active users and estimate 30% would use the feature, Reach = 3,000. Avoid percentage estimates directly in the Reach column: use absolute numbers so scores are comparable across features.
Define your Impact scale
Impact uses a fixed multiplier scale, not a freeform number. In cell C2, add a dropdown validation (Data > Data Validation > List): 3, 2, 1, 0.5, 0.25. Define each level in a reference tab: 3 = multiple orders of magnitude improvement to core metric, 2 = significant improvement, 1 = moderate improvement, 0.5 = minimal improvement, 0.25 = barely noticeable. This reference tab prevents score inflation — the most common calibration problem.
Add the Confidence column with honest estimates
Confidence is the most commonly gamed column. Teams always set it at 100% to inflate scores. Add a Confidence guide in the reference tab: 100% = you have A/B test data or strong user research supporting the claim. 80% = you have indirect data or a few interviews. 50% = you have a hypothesis with no direct evidence. 20% = pure assumption. Use percentage values in the cell (80%, not 0.8) and make sure the formula handles the conversion.
Enter Effort in person-months and write the RICE formula
Effort is total person-months across all roles: engineering + design + PM. A feature requiring 2 engineers for 1 month plus 0.5 designer months = 2.5 person-months. In the RICE Score column, enter: =(B2*C2*D2)/E2 where B=Reach, C=Impact, D=Confidence (as decimal), E=Effort. Format the score column as a number with no decimal places. Add conditional formatting: green for top quartile, yellow for middle, red for bottom.
Sort and use the output for roadmap decisions
Sort the sheet by RICE Score descending. The top items are your highest-priority candidates. Do not blindly build the top 10 — use the RICE score as a starting point, then apply strategic filters: Are these items aligned with the current quarter's goal? Do they have dependencies that require reordering? Are there items with low RICE but high strategic importance (executive mandate, compliance requirement)? The score informs judgment; it does not replace it.
Common mistakes
Scoring in isolation instead of as a team
When one person fills in all the RICE scores, the model reflects one perspective. Have engineering estimate Effort independently from the PM's Reach and Impact estimates. Bring the scores together and discuss disagreements. The conversation is often more valuable than the final score.
Setting confidence too high by default
The average RICE model has confidence of 80% or higher on every feature because no one wants to admit uncertainty. Cap the average confidence of your backlog at 60%. If it is higher, you have not been honest about what you know.
Comparing RICE scores across different product areas
A feature touching the onboarding flow has high Reach by definition (every new user). A feature for power users has lower Reach but may have higher Impact per user. RICE scores favor broad, shallow features over deep, narrow ones. Segment your backlog by product area and score within segments.
Ignoring the denominator (Effort)
Teams focus on maximizing the numerator (Reach × Impact × Confidence) and accept high Effort as a constant. Effort is a lever: an MVP version of a feature with half the Effort often scores twice as high. Ask: "What is the smallest version that earns most of the Reach and Impact?" before accepting the full-effort estimate.
Tips
Add a "Last Scored" date column so you know which scores are stale — rescore quarterly
Create a second tab for "Rejected Features" with a required "Rejection Reason" column — this prevents rejected features from reappearing in planning without context
Use Google Sheets conditional formatting to highlight Effort > 3 person-months in orange as a prompt to explore MVPs
Share the scoring sheet with engineering leads before sprint planning so everyone arrives with the same prioritization data
How Vantage helps
Vantage extracts requirements from PRDs and assigns effort estimates based on the generated tickets. You can export these into a RICE scoring spreadsheet with Effort pre-populated from the ticket generation output, reducing the manual estimation step that consumes hours of engineering time in planning sessions.