RICE is a prioritisation method that gives each backlog item a score from four estimates: reach (how many people the item affects in a period), impact (how much it affects each one), confidence (how much the estimates rest on evidence) and effort (how much work it costs). The formula is reach × impact × confidence ÷ effort. The score does not decide on its own, but it makes the assumptions behind each bet explicit.
How to use
- Choose the reach period (month, quarter or year) and use the same one for every item.
- Write one item per row and estimate reach, impact, confidence and effort.
- Watch the table re-sort by score, with the top three highlighted.
- Read the warnings: empty effort zeroes the score and 50% confidence signals a hunch.
- Export as PNG, CSV or Markdown, or create a link to review with the team.
What each field asks for
| Field | What to enter |
|---|---|
| Reach period | Every item must use the same period for reach. |
| Item | E.g. export report as PDF |
| Reach (R) | People or events affected in the period. |
| Impact (I) | How much each person reached is affected. |
| Confidence (C) | How much the estimates rest on evidence. |
| Effort (E) | Person-months for the whole team. |
Example
A finance app team compares four items for the quarter. The payment reminder by email reaches 1,200 people, has impact 2, 80% confidence and costs 1.5 person-months: 1,200 × 2 × 0.8 ÷ 1.5 = 1,280 points, first place. The PDF report reaches more people (2,000) but costs 4 person-months and scores 800. Dark mode has the largest reach, yet it is a hunch (50% confidence) with low impact: 375 points.
When to use
- When the backlog has more good items than capacity and the debate has turned into a clash of opinions.
- Before quarterly planning, to bring the team an order with explicit assumptions.
- To compare items of different kinds (improvement, fix, new feature) on the same scale.
- To revisit an old decision: changing one estimate shows at once whether the order changes.
Frequently asked questions
What is the RICE score?
It is reach × impact × confidence ÷ effort. The higher it is, the more expected value per unit of work. It ranks comparable items; it does not measure absolute value.
How do I estimate reach?
Count people or events affected in a fixed period, such as customers who use the screen per quarter. Use usage data where you have it; without data, a conservative estimate and a lower confidence.
Which impact scale should I use?
The usual scale is 3 (massive), 2 (high), 1 (medium), 0.5 (low) and 0.25 (minimal). What matters is that the team uses the same scale and the same definition of impact for every item.
What if confidence is low?
A 50% confidence signals a hunch. Instead of building the item, plan a cheap test that adds evidence, such as an interview or a prototype, and score it again.
What is the difference between RICE and ICE?
ICE uses impact, confidence and ease, without reach. It is faster, but it mixes audience size with impact. RICE separates the two and works better when reach varies a lot between items.
What are the limits of RICE?
It does not capture dependencies, technical risk, regulatory deadlines or strategic bets. Use the score to open the discussion and record why the final order differs from the table, if it does.
How should I present the result to the team?
Show the table with the assumptions for each item, not just the score. Export a PNG for the presentation or create a link: anyone who opens it can adjust an estimate and see the effect on the order.