A RICE Recipe Your Team Will Actually Use

You have read about RICE. Maybe you even pasted the formula into a doc. But your team has not scored a single project with it, and the roadmap still gets set by whoever argued loudest in the last meeting. That is the gap. RICE is not hard to understand. It is hard to run for the first time, because the first session is where people freeze on the numbers, fight over impact, and decide the whole thing is fake math.

This brief hands you the recipe and points at the spots where first attempts fall apart. The goal is a scored backlog by Friday, not a perfect one.

The deep cut

• RICE turns opinions into numbers you can argue with. Notion frames it as a way to silence the loudest gut and rank projects on Reach, Impact, Confidence, Effort.
• A backlog nobody scored is just a wish list. Teams read about RICE and never touch real data, so the CEO's pet feature still jumps the line.
• Score two real items this week, not the whole backlog. Waleed Elaghil pushes the hands-on session over more reading.

What the four letters actually mean

RICE stands for Reach, Impact, Confidence, and Effort. Reach is how many people the project touches in a set window, say a quarter. Impact is how much it moves them, scored on a fixed scale from 3 for massive down to 0.25 for minimal. Confidence is how sure you are about your own guesses, as a percent. Effort is the work in person-months. You divide the first three by Effort, and the bigger number wins.

The point is not the arithmetic. As Notion puts it, RICE turns abstract concepts and intangibles into quantifiable data points. Three of the four scales are fixed menus you pick from, which is on purpose. It stops people from stalling on whether impact is a 1.7 or a 1.8.

Score real items, not made-up ones

The fastest way to make RICE stick is to run it on things your team already wants to build. Joe Cotellese does this with his own app, ClipDish, and it is worth copying because he shows the reasoning out loud. Auto-categorizing recipes reaches everyone who saves one, about 2000 people a quarter, and he calls the impact massive because customers keep asking for it.

Compare that to a currency feature for non-US users. He looked at analytics and found only 50 such people a month, so both Reach and Impact drop hard. Same effort, very different score. That is the whole method working: two features that felt equally reasonable in a meeting land in different spots once you use real numbers. Have your team pick two features and talk through them the same way.

The traps that kill a first attempt

Three things wreck first sessions. First, treating the score as law. Notion is blunt that RICE is a guide, not the law, and the numbers are guesstimates you re-check as things change. If a project with a lower score is clearly strategic, ship it anyway and move on.

Second, letting engineers freeze on Effort. Cotellese notes that engineers avoid ambiguity and hate committing to a number. Tell them plainly this is not a deadline, just a rough size next to other items. Third, scoring Impact in a vacuum. He raises a real case: a non-US feature scores low until you fund a non-US market push, and then the impact number goes up. Context changes the score, so name your goal before you score.

Make it a habit, not a one-off

A single scoring session does not stick. Pick a cadence and re-score, because your estimates were guesses and the market moves. Elaghil frames the real question for a team as how often to re-score, and the answer is tied to your planning rhythm, usually each quarter.

Keep it in a shared sheet or a template so the scores live somewhere everyone can see and question. Notion offers a worksheet for this, and any spreadsheet works fine. If your team learns better by watching, the Airfocus walkthrough shows the formula run on live examples. The tool matters less than the ritual of doing it out loud, together, on things that are real.

Three questions for your team

• What single goal are we scoring against this quarter? Pick one, like revenue or reach, so Impact scores mean the same thing to everyone.
• Which two backlog items do we score first, out loud, together? Start with features you are already debating so people see the math change the answer.
• How often do we re-score, and who owns the sheet? Tie it to your planning cadence so the numbers stay honest instead of going stale.