Strategy
Rank growth ideas by Impact, Confidence and Ease, and kill the noise.
- Category
- Strategy
- Signup
- None
- Your data
- Stays in your browser
- Also in
- Claude, via MCP
How do you calculate an ICE score?
Rate each idea from 1 to 10 for Impact (how much it could move the goal), Confidence (how sure you are it will work) and Ease (how little effort it takes). ICE score = (Impact + Confidence + Ease) ÷ 3. Rank ideas by score and start at the top.
Your highest-ICE bet is Add live chat to pricing page at 7.3/10. Run it next. ICE keeps you honest, a flashy high-impact idea with low confidence and ease loses to a boring, sure, easy win more often than founders expect.
Score each 1–10. ICE is deliberately blunt, its job is to force a ranked order and end debate, not to be precise. Re-score as you learn.
About this calculator
Every growth team has a backlog longer than its capacity, and the loudest idea in the room usually isn't the best one. ICE scoring forces a blunt, defensible ranking so debate ends and execution starts, rating each idea on Impact, Confidence, and Ease strips away the persuasiveness of whoever's pitching and leaves you with a number.
How to use it
- Add each growth idea by name.
- Score Impact, Confidence, and Ease from 1 to 10 for each: Impact is how much it moves the needle if it works, Confidence is how sure you are it will work, Ease is how quickly and cheaply it can ship.
- Read the ranked list, sorted highest ICE score first, with the top idea called out as the next bet.
Methodology
ICE score is the simple average of the three inputs for each idea: (Impact + Confidence + Ease) ÷ 3, giving a score out of 10.
Ideas are ranked in descending order of ICE score, the highest-scoring idea is surfaced as the recommended next bet.
The framework is intentionally blunt, it's designed to force a ranked order and end prioritization debates, not to precisely quantify expected value. A flashy high-impact idea that scores low on confidence and ease will lose to a boring, sure, cheap win more often than most teams expect, and that's the point.
Scores are subjective inputs from your team, not measured data, so ICE is only as good as the honesty of the ratings. Re-score as you learn more about what actually works.
ICE score
ICE score = (Impact + Confidence + Ease) ÷ 3 Rank ideas from highest to lowest score
- Impact
- How much the idea could move the metric you are focused on, if it works.
- Confidence
- How strong the evidence is that it will work: data, past tests or customer feedback.
- Ease
- How little time, money and coordination it needs. 10 is very easy.
ICE is deliberately rough. Its value is in forcing a ranked list and a shared conversation about why each idea scored as it did. A related method, RICE, adds Reach and divides by effort; Intercom describes it in its RICE prioritisation guide. Use RICE when ideas differ a lot in how many people they affect.
Worked example: the three default ideas
A growth team is choosing between a partner referral programme (Impact 8, Confidence 6, Ease 7), rebuilding onboarding (9, 7, 4) and adding live chat to the pricing page (5, 8, 9).
- Referral programme = (8 + 6 + 7) ÷ 3 = 7.0.
- Onboarding rebuild = (9 + 7 + 4) ÷ 3 = 6.7.
- Live chat on pricing = (5 + 8 + 9) ÷ 3 = 7.3.
- Ranking: live chat (7.3), referral programme (7.0), onboarding rebuild (6.7).
The quick, likely win goes first, and the big onboarding project waits despite the highest impact score. That is often right, but check it: if onboarding is where most customers are lost, its impact may be underrated. Size the effect on the funnel with the conversion funnel calculator, then re-score after the first test.
How to score ideas well
ICE works when scoring is consistent:
- Define the goal first. Impact means nothing without one metric, such as qualified leads or activation rate.
- Write down why. A one-line reason for each score stops the same idea being scored differently next week.
- Score confidence on evidence, not enthusiasm: data and past tests earn high scores, opinions earn low ones.
- Re-score after each test, because results change confidence for related ideas.
Treat close scores as ties and decide on strategic fit.
Prioritising growth ideas in India vs the US
Indian teams often have more people than budget, so Ease should weigh money more than effort. Ideas using WhatsApp, regional language content or partnerships can score high on ease and impact for Indian audiences. Check for festive-season timing, which can make an easy idea hard to test in October and November.
US teams often have more budget than people, so Ease should weigh engineering and design time heavily. Many US growth teams move to RICE once they have reliable reach data from analytics. Pair prioritisation with a testing plan using the A/B test significance calculator.
Common ICE scoring mistakes
- Scoring without a single goal, so impact means different things for each idea.
- Letting the most senior person's scores count most.
- Giving high confidence to ideas with no evidence.
- Never re-scoring, so the list goes stale.
- Using ICE for large strategic choices that need proper analysis.
Sources
- RICE: Simple prioritization for product managers, Intercom. Accessed 2026-10-19.
FAQ
Because ICE averages all three factors equally. A huge-impact idea with low confidence and high effort (say, a 9/3/2) scores 4.7, while a modest-impact, high-confidence, easy win (say, 5/8/9) scores 7.3. ICE is designed to surface exactly that kind of trade-off, big swings that are unlikely to land or expensive to test lose to smaller sure bets.
Ideally the people closest to execution and the data, a marketer scoring their own idea's impact tends to inflate it. Having a second person or the whole team score independently, then averaging or discussing outliers, produces more honest numbers than letting each idea's champion self-score.
RICE adds a Reach factor (how many users or customers it touches) and divides by Effort instead of averaging Ease, making it more granular for product prioritization at scale. ICE is deliberately simpler and faster, better suited to a smaller backlog of growth experiments where speed of decision matters more than precision.
Yes, that's the whole value of the framework. A launched idea should get its Confidence score updated based on real results, and remaining backlog items should be re-ranked against what you've just learned. ICE is a living prioritization tool, not a one-time ranking.
ICE is widely credited to Sean Ellis and the growth hacking community, as a quick way to rank experiments.
In standard ICE, yes. Some teams weight impact more. If you do, keep the weights fixed so scores stay comparable.
Enough to choose the next few weeks of work, often ten to twenty. Very long lists get scored carelessly.
Numbers looking off?
If the model says one thing and the pipeline says another, that gap is usually where the revenue is. A 30-minute call finds it.
It starts with a 30-minute call. Pick a time below.