The RICE scoring model is a prioritization framework that ranks projects or features by a single number built from four factors: reach, impact, confidence, and effort. You multiply reach by impact by confidence, divide by effort, and the result is a comparable score you can sort a backlog by. Higher scores go first. RICE was created at Intercom to stop the loudest idea from winning and force every candidate through the same math.
Key takeaways
- RICE stands for Reach, Impact, Confidence, and Effort. The score is (Reach x Impact x Confidence) / Effort, which reads as total impact per unit of work.
- Reach is people or events per time period, impact uses a fixed scale (0.25 to 3), confidence is a percentage, and effort is person-months.
- RICE suits data-rich product backlogs where you can estimate reach honestly. Its blind spot is time: it has no urgency term, so a deadline-driven item can score too low.
- For funding whole portfolios rather than sequencing one backlog, a weighted scoring model against strategy and capacity usually fits better.
What is the RICE scoring model?
The RICE scoring model is a way to prioritize a list of projects, features, or ideas by scoring each one on four factors and combining them into one number. It was developed at Intercom, a customer messaging company, to make roadmap decisions less about who argued hardest and more about expected value per unit of effort. Each candidate gets a RICE score, the list is sorted from high to low, and the top of the list is what you work on first.
What makes RICE useful is that it is deliberately simple. Four factors, one formula, one number. It does not try to model every nuance of a decision. It gives a product or portfolio team a fast, defensible first ranking that anyone can reproduce, which is usually enough to settle the easy calls and focus the argument on the genuinely close ones.
What does RICE stand for?
RICE is an acronym for the four factors you score. The first three raise the score because more is better; effort lowers it because more work for the same return is worse. The table lays out what each one measures and the unit it is scored in.
| Letter | Factor | What it measures | Typical unit |
|---|---|---|---|
| R | Reach | How many people or events the item affects in a set period | Users or events per quarter |
| I | Impact | How much it moves the needle for each person reached | Fixed scale: 3, 2, 1, 0.5, 0.25 |
| C | Confidence | How sure you are of your reach and impact estimates | Percentage: 100%, 80%, 50% |
| E | Effort | Total work required to deliver the item | Person-months |
What is the RICE formula?
The RICE formula is (Reach x Impact x Confidence) / Effort. You multiply the three positive factors together, then divide by effort. The numerator captures how much total good the item does, adjusted for how sure you are; dividing by effort turns it into value per unit of work, which is what you want to maximize when capacity is limited.
RICE score = (Reach x Impact x Confidence) / Effort
Because confidence is a percentage, use it as a decimal in the math (80 percent becomes 0.8). The output is not a meaningful quantity on its own; a score of 2,000 does not mean anything in isolation. It only matters relative to the other items scored the same way, which is why you must score every candidate with the same definitions.
How do you score each RICE factor?
The score is only as good as the inputs, and the four factors each have a standard way to estimate them. Score them in order, and write your assumptions down so the numbers can be challenged later.
Reach
Reach is how many people or events the item touches in a defined time window, such as customers per quarter or transactions per month. Use real numbers from analytics wherever you can rather than a gut feel. Pick one time period and hold it constant across every item, or the scores will not be comparable.
Impact
Impact is how much the item moves your goal for each person reached, scored on a fixed scale: 3 for massive, 2 for high, 1 for medium, 0.5 for low, and 0.25 for minimal. The scale is intentionally coarse. You are not measuring impact precisely, you are agreeing whether an item is a big deal or a small one, and the multipliers keep those judgments consistent.
Confidence
Confidence is a percentage that discounts the score for how much evidence you actually have. Intercom uses 100 percent for high confidence backed by data, 80 percent for medium, and 50 percent for low or mostly a hunch. Confidence is the brake on exciting but unproven ideas: a project with huge claimed reach and impact but no data behind it gets multiplied down until it earns real evidence.
Effort
Effort is the total work to deliver, estimated in person-months, counting design, engineering, and any other function involved. It is the one factor that divides rather than multiplies, so a cheap item with modest impact can still beat an expensive one. Estimate effort at the same rough level as everything else; do not agonize over precision the model does not need.
RICE worked example
Say a product team is choosing between three features. Scoring each on the four factors and running the formula turns a debate into a ranked list.
| Feature | Reach | Impact | Confidence | Effort | RICE score |
|---|---|---|---|---|---|
| Onboarding redesign | 5,000 | 2 | 0.8 | 4 | 2,000 |
| Billing export | 1,200 | 1 | 1.0 | 1 | 1,200 |
| New dashboard widget | 8,000 | 0.5 | 0.5 | 3 | 667 |
The onboarding redesign wins at 2,000, calculated as (5,000 x 2 x 0.8) / 4. The dashboard widget reaches the most people but scores lowest, because its impact per person is minimal and the team is only half sure of the estimate. That is RICE doing its job: it stops a big reach number from carrying a weak feature to the top of the list.
What is a good RICE score?
There is no universal good RICE score. The number is only meaningful relative to the other items scored with the same definitions, so a score of 500 might top one backlog and sit at the bottom of another. What matters is the ranking, not the absolute value. Trying to compare RICE scores across teams that scored reach, impact, or effort differently is the most common way the model gets misused.
Use the score to order the list and to see how close the top items are. When the top two or three scores are far apart, the decision is easy. When they are within a few percent of each other, treat that as a signal that RICE cannot separate them and the call needs judgment, capacity data, or a second factor RICE does not model, such as a hard deadline.
What is the difference between RICE and WSJF?
RICE ranks by impact per unit of effort and has no time dimension, while WSJF (weighted shortest job first) ranks by economic urgency, dividing cost of delay by job size so time-critical work rises. RICE fits data-rich product backlogs; WSJF fits agile and SAFe portfolios where sequencing across teams and deadlines is the real problem. The practical difference is that WSJF captures urgency where RICE cannot.
If a compliance date or a closing market window makes a low-reach item urgent, RICE will underrate it because none of its four factors sees time, whereas WSJF has a time-criticality term built in. Many teams keep both on the shelf and pick per decision: RICE when the question is value per effort, WSJF when the question is what loses the most value by waiting. For the economics underneath urgency, see cost of delay.
What are the pros and cons of RICE?
RICE is popular because it is quick, transparent, and hard to game once the definitions are fixed. It is not a complete decision system, and knowing where it falls short keeps you from trusting it past its range.
| Strengths | Limitations |
|---|---|
| Fast to apply and easy to explain | No time or urgency dimension |
| Confidence factor curbs unproven hype | Reach and impact estimates can be soft without good data |
| Produces a reproducible ranked list | Ignores dependencies and strategic fit |
| Rewards low-effort, high-value work | Scores are not comparable across differently-scored lists |
When should you use RICE?
Use RICE when you are sequencing a backlog of features or improvements within one product and you have at least rough data on how many people each item reaches. It is at its best on data-rich product roadmaps where reach is a real number, not a guess. It is weakest for comparing unlike investments across a whole portfolio, where strategic fit, risk, and capacity matter more than impact per person.
For portfolio-level decisions that weigh strategy, return, and risk against limited funding, a weighted multi-criteria approach fits better than RICE. See the project scoring model for how RICE, WSJF, and weighted scoring compare, the prioritization frameworks guide for choosing among them, and how to prioritize a project portfolio for scoring against real capacity. To sort a long list fast before scoring, start with the impact effort matrix.