There will always be more sensible-looking opportunities than a team can take on. Some are big and distant. Some are small but immediately useful. Some reduce risk. Some are simply things that customers have been asking for loudly. The job is to make a decision that people can understand, even when they would have made a different one.
RICE is one way to do that. It stands for Reach, Impact, Confidence and Effort. The formula is simple: Reach × Impact × Confidence / Effort. The numbers are less important than the conversation behind them.
Start with the horizon
Before scoring anything, agree the horizon you are working to. A five-year bet might be exactly right for a company that is willing and able to wait for it. It is probably the wrong bet if the vision you are trying to achieve is only three years away.
The horizon anchors the rest of the conversation. It says how much uncertainty you can afford to take on, how long you can wait for a return, and what kind of risk appetite you have. A small opportunity with a short horizon can be more valuable than a large opportunity that arrives too late to matter.
Reach and impact
Reach is how many people, accounts or transactions will actually benefit in the period you are considering. Be honest about the period. “Ten thousand users eventually” is not the same as two thousand active users this quarter.
Impact is how much better it will be for each of those people. If you can put a real number on it, great. If you cannot, use a consistent scale. The point is to compare opportunities, not to make the spreadsheet look scientific.
Confidence and effort
Confidence is where you make the uncertainty visible. A large opportunity with weak evidence might still be worth pursuing, but it should be recognised as a bet. If confidence is too low, the next step is often learning, not delivery.
Effort is the real cost, not just engineering time. It includes design, research, testing, rollout, support and the cost of keeping the thing running afterwards. A rough shared unit, such as person-weeks, is usually enough early on.
Be careful with the maths
RICE gets less useful when the inputs are guesses on top of guesses. Making assumptions is normal. The problem is that multiplying several shaky assumptions together can give you a very precise-looking number that means very little.
That does not mean you should abandon the framework. It means you should say what is known, what is assumed and where the number is most sensitive. Use the score to start a conversation, not to end one.
A small example
Imagine choosing between improving password reset and building a reporting dashboard. Password reset reaches 12,000 users a month, has good evidence behind it and takes two person-weeks. The dashboard reaches 800 operations users, could have a large impact, but will take eight person-weeks and has more uncertainty around it.
RICE may put password reset first. That does not say the dashboard is a bad idea. It rationalises the sequencing: do the smaller, proven piece now, then decide whether the larger bet is still the right next step.
A score should never overrule something genuinely non-negotiable, such as a security issue, a regulatory commitment or a major strategic move. It is there to make the decision clearer, not to make it for you.