What is ICE Scoring?

ICE scoring, in depth

ICE optimizes for throughput: a backlog of thirty items can be scored in one sitting, which makes it popular for growth experiments and early-stage backlogs where estimating reach precisely is impossible anyway. The three 1–10 gut ratings are acknowledged approximations.

That looseness is also the caveat — ICE scores drift with mood and scorer. It works best as a first-pass sort that queues deeper discussion for the top of the list, and it works better with several scorers averaging than with one enthusiast rating their own ideas.

Putting it into practice

Score quickly, sort, then sanity-check the top five against actual demand before committing. If a high-ICE idea has zero presence in your feedback inbox or roadmap votes, that absence is information — either an awareness gap or a confidence score that was really hope.

Related terms

• RICE scoring — RICE scores features by Reach × Impact × Confidence ÷ Effort — how many users it touches, how much it moves them, how sure you are, and what it costs to build.
• Prioritization framework — A prioritization framework is a structured method for deciding what to build next — RICE, ICE, MoSCoW, impact–effort matrices — that scores or sorts candidates on explicit criteria instead of instinct alone..
• Impact–effort matrix — An impact–effort matrix plots candidates on two axes — value delivered versus work required — yielding four quadrants: quick wins, big bets, fill-ins, and money pits.