RICE scoring explained (with a simple template)

RICE is a scoring framework for comparing initiatives that have nothing else in common — a backend migration against a new onboarding flow against a pricing experiment. It doesn't tell you what's right. It gives everyone the same four questions to answer, so the comparison is at least honest about what it's trading off.

The name is an acronym for the four inputs: Reach, Impact, Confidence, and Effort. Score each one, run them through a simple formula, and you get a number you can sort a backlog by. The value isn't in the precision of the number — it's in forcing the conversation about each input before you commit.

The four inputs

Input	What it measures	Typical scale
Reach	How many people this touches in a given period	A real number — "reach per quarter," e.g. 1,200 users
Impact	How much it moves the needle for each person it reaches	Usually 0.25 (minimal) to 3 (massive), on a fixed scale
Confidence	How sure you are about your Reach and Impact estimates	Percentage — 100%, 80%, or 50%
Effort	Total person-months to build it	A real number — 0.5, 1, 3, and so on

The formula combines them into a single score:

Higher scores mean more expected value per unit of effort. That's it — the framework is deliberately simple. Most of the real work happens in how honestly you fill in the four inputs, not in the arithmetic.

A worked example

Say you're comparing two initiatives for next quarter:

• Saved filters — reaches an estimated 2,000 users per quarter, with a modest impact score of 1 (noticeable but not transformative), 80% confidence since you have some usage data backing the estimate, and 1 person-month of effort. Score: (2,000 × 1 × 0.8) ÷ 1 = 1,600.
• Onboarding redesign — reaches all 500 new users per quarter, with a high impact score of 3 (directly tied to activation), but only 50% confidence since it's based on a hypothesis rather than data, and 3 person-months of effort. Score: (500 × 3 × 0.5) ÷ 3 = 250.

On paper, saved filters wins by a wide margin — not because it's the more exciting idea, but because it reaches far more people for far less effort, even with a smaller per-person impact. That's the kind of trade-off RICE is good at surfacing, and the kind that's easy to miss when you're prioritising by gut feel or by whoever pitched most recently.

Where RICE breaks down

A few things to watch for:

• Confidence becomes a fudge factor. It's tempting to nudge confidence up on ideas you already like and down on ones you don't. Treat it as a genuine measure of evidence quality — do you have data, or are you guessing — not a knob for steering the outcome.
• Reach gets inflated. "Everyone will see this" is rarely true. Use the actual number of people affected in the period you're scoring against, not the theoretical total user base.
• Effort excludes the unglamorous parts. QA, migration, documentation, and support load all count. An estimate that only covers build time will consistently favour initiatives that look cheaper than they are.
• It compares initiatives, not values. RICE can tell you saved filters beats onboarding redesign on expected value per effort. It can't tell you whether activation matters more to the business right now than engagement — that's still a judgement call, and it's supposed to be yours.

A simple template to copy

For each initiative, fill in:

1. Reach — number of people affected per quarter, with the source of that estimate
2. Impact — 0.25 / 0.5 / 1 / 2 / 3, with a one-line reason for the score
3. Confidence — 50% / 80% / 100%, based on how much real evidence backs Reach and Impact
4. Effort — person-months, including QA, docs, and rollout
5. Score — (Reach × Impact × Confidence) ÷ Effort

Run every initiative you're comparing through the same five lines, sort by score, and use the ranking as a starting point for the conversation — not the final word. The framework's job is to make the trade-offs legible. Deciding what to do about them is still yours.