DTC growth teams running paid acquisition, CRO, and lifecycle in the same sprint hit the same problem every planning cycle: more reasonable ideas than the team can ship, and no honest way to compare a landing-page test against a lifecycle email rebuild against a paid-creative refresh. RICE prioritization is the most widely adopted framework that forces those comparisons onto a single number — a scoring system that ranks a backlog of initiatives by expected value per unit of effort.
Reach: count of customers/sessions per period. Impact: 3 / 2 / 1 / 0.5 / 0.25. Confidence: 100% / 80% / 50%. Effort: person-months.
Each input has a defined shape. Reach is a concrete count of customers, sessions, or events affected per period (per month is the common choice), not a rating. Impact is a categorical multiplier on a standard scale: 3 for massive, 2 for high, 1 for medium, 0.5 for low, 0.25 for minimal. Confidence is a percentage anchored at 100%, 80%, and 50%, reflecting how sure the estimator is of the Reach and Impact numbers, not how excited they are about the idea. Effort is denominated in person-months — one person working full-time for one month. The output is a single number; higher means ship-sooner.
What RICE is genuinely good at is making the implicit explicit. The Confidence multiplier forces estimators to declare uncertainty out loud rather than bury it in a vague Impact estimate or padded Effort. On a backlog of A/B tests and other CRO bets, that discipline alone is worth the exercise. The team-level argument about the inputs is where most of the value lives; RICE done solo produces a sorted list with no calibration.
Where RICE breaks is across very different timescales and across strategic optionality. A six-month brand campaign and a two-week landing-page test will almost never score comparably, because Effort dominates the denominator and the small test usually wins the math even when the campaign is the right call. The framework also doesn’t capture optionality: a low-RICE test whose real value is the much larger follow-up it unlocks has no slot in the formula. The same blind spot affects incrementality work, where payoff comes from sharper measurement rather than tracked direct lift.
ICE — Reach dropped from the formula — is the lighter-weight cousin used by growth teams running rapid experimentation cycles where most tests reach the same audience and the Reach column adds noise without signal. RICE is the right choice when initiatives have meaningfully different reach profiles. Either way, the operator move is the same: score as a team, argue the inputs, and treat the final number as a tiebreaker among already-defensible bets rather than a verdict.