RICE scoring: the prioritization formula, with worked examples

RICE scoring ranks ideas by a single number: reach times impact times confidence, divided by effort. You score each of the four factors separately, run the formula, and sort the list. The point is not that the number is precise, because it is built from estimates, but that scoring all four factors on their own stops one loud advocate from winning a priority fight on enthusiasm alone.

The model comes from Intercom, which built it to compare feature ideas that all sounded important. It has spread because it is quick, it is defensible, and it forces the two questions people most like to skip: how sure are you, and how much work is this really. This article covers what each factor means, the formula, a worked example, sensible scales, and the places RICE quietly misleads a team.

What is RICE scoring?

RICE scoring is a prioritization method that assigns each idea a score from four inputs: Reach, Impact, Confidence and Effort. Reach is how many people or events the idea affects in a set period. Impact is how much it moves your goal for each of them. Confidence is how sure you are of the reach and impact numbers, expressed as a percentage. Effort is the total work in person-months. You multiply the first three, divide by the fourth, and the higher the result the higher the priority.

The design does one clever thing: it puts confidence directly into the math. Most prioritization schemes let an exciting-sounding idea ride on a wildly optimistic guess. RICE makes you write down how solid the guess is, and a low confidence percentage drags an otherwise flashy score back down to earth. That single discipline is why teams reach for RICE when the list of candidates is long and everyone has a favorite. It is one of the more rigorous tools in idea prioritization precisely because it refuses to treat a hunch and a measured result as equal.

How do you calculate a RICE score?

The formula is Reach times Impact times Confidence, all divided by Effort. Score each factor on its own before you combine them, using the same scale for every idea so the comparison is fair. The output is a score in units that mean nothing in isolation and everything in relation to the other ideas on the list, so RICE is always a ranking exercise, never a single verdict.

Work an example. Say a checkout redesign would reach 4,000 shoppers a quarter, you judge its impact as high (2), you are 80% confident in those figures, and it will take 4 person-months. The score is 4,000 times 2 times 0.8 divided by 4, which is 1,600. A second idea, a referral prompt, reaches 8,000 people at medium impact (1) with 50% confidence and 2 person-months of work: 8,000 times 1 times 0.5 divided by 2, which is 2,000. The referral prompt wins, even though the redesign felt bigger, because it touches more people for less work and RICE made that trade explicit.

What is a good RICE score?

There is no universal good RICE score, because the number depends entirely on how you set your reach counts and effort units. A score of 2,000 is only meaningful next to the other scores on the same list, scored by the same person with the same scales. Comparing a RICE score from one team to another, or from one quarter to a differently-calibrated quarter, tells you nothing. Treat the score as a sorting key, not a grade.

What matters more than the absolute figure is consistency. Score every idea in one sitting, with one definition of Impact and one time window for Reach, so the ranking reflects real differences between ideas rather than drift in how you were counting. A score is only as trustworthy as the numbers under it, so it pays to trust the data behind each estimate before you lean on the ranking: if your reach figure comes from a stale dashboard or a metric nobody owns, the confidence percentage should reflect that, not paper over it.

RICE vs the impact effort matrix: which should you use?

Use the impact effort matrix for a fast, visual first cut and RICE when you need a defensible ranking of a longer list. The matrix plots ideas on two axes and sorts them into four quadrants in about ten minutes, which is perfect for a live workshop. RICE takes longer because it adds reach and confidence and produces an ordered list rather than four buckets, which is what you want when a roadmap decision needs to survive scrutiny.

Many teams run both in sequence: sketch an impact effort matrix to throw out the obvious thankless tasks, then RICE-score the survivors to rank the contenders. The matrix answers "is this worth considering at all," and RICE answers "in what order." Neither replaces judgment; both exist to make the judgment visible and harder to fudge.

What are the limitations of RICE scoring?

The core limitation is false precision. A RICE score of 1,847 looks authoritative, but it is a product of four estimates, three of which are often guesses, and multiplying guesses multiplies their error. Two people scoring the same idea can land 3x apart, and the formula gives no error bars. The number organizes the argument; it does not resolve the uncertainty, and a team that forgets this ends up defending decimal places that were never real.

The second limitation is that RICE flattens strategy into throughput. It rewards ideas that touch many people cheaply, which quietly biases a roadmap toward incremental wins and against the rare, expensive, low-reach bet that defines a product. It also ignores dependencies, timing and sequencing entirely: a low-scoring item might be a prerequisite for a high-scoring one. Use RICE to rank comparable candidates, then override it deliberately when strategy demands, and write down why. The score is an input to the decision, not the decision.

The short version

RICE scoring multiplies reach, impact and confidence and divides by effort to rank ideas, and its real value is forcing confidence and effort into a conversation that usually skips them. Score every factor on a fixed scale in one sitting, treat the result as a ranking rather than a grade, and keep the estimates honest by grounding reach in data you trust. Then use the ranking as a starting point for judgment, not a substitute for it. The best way to feed a RICE model is with a strong, varied set of candidates in the first place, which is where structured ideation earns its keep.