ICE Score Calculator
You will always have more ideas than sprints. ICE is the fastest honest way to rank them: rate each idea on Impact, Confidence and Ease, and let the average sort your backlog so the best bets rise to the top.
ICE score = the average of three 1–10 ratings — Impact (how much it moves the metric), Confidence (how sure you are it will), and Ease (how cheap and fast it is). Popularized by Sean Ellis for growth-experiment prioritization, ICE deliberately trades precision for speed: it is a gut-check ranking, not a forecast. Score every idea the same way, sort high to low, and work the list from the top.
ICE Score Calculator inputs and result
| ICE average | Verdict | Why |
|---|
How to use this calculator
- Score Impact 1 to 10Estimate how much the idea would move your target metric if it worked. Anchor the extremes: a 1 is a rounding error, a 10 is a step-change in the number that matters.
- Score Confidence 1 to 10Rate how much evidence stands behind that Impact guess. Past experiments, real data, or a strong analogue earn a high score; a hallway hunch earns a low one. This axis is what keeps ICE honest.
- Score Ease 1 to 10Rate how cheap and fast the idea is to ship. A higher score means less effort. Score Ease, not difficulty, so all three axes point the same way — higher is always better.
- Read the average and the weakest axisThe tool averages the three and gives a verdict band. It also names your lowest axis — the single thing to fix to move the idea up the list.
- Score every idea and sort the listICE only works as a ranking. Score the whole backlog the same way, sort high to low, and work the top. Export the CSV to build your prioritized roadmap.
RGM Expert Says
We reach for ICE when a team is drowning in ideas and starved of decisions. The whole point of the framework is speed: you are not building a financial model, you are forcing three quick, comparable judgments so the backlog stops being a popularity contest. The discipline is in scoring every idea on the same 1–10 scale in one sitting, with the same people, so the numbers mean the same thing across the list.
Confidence is the axis that earns ICE its keep, and the one teams cheat on. It is easy to fall in love with a high-Impact idea and quietly inflate how sure you are it will work. We force the question ‘what evidence makes you confident?’ for every score above a six. If the only answer is enthusiasm, the Confidence drops, the ICE drops, and the idea waits behind something better-supported. That single habit kills more bad bets than any amount of analysis.
Where ICE stops being enough is when ideas differ wildly in cost or reach. A one-day tweak and a two-month rebuild can post the same ICE, which is misleading. That is the moment we graduate the team to RICE, which divides by effort and adds reach, so big-but-expensive ideas are compared fairly. ICE gets you moving; RICE keeps you honest at scale.
How it works
ICE is intentionally simple: three ratings on the same scale, averaged into one comparable number. We show both the average (the common convention) and the product I×C×E, which some teams prefer because it spreads scores apart.
- Impact — how much the idea moves your target metric, 1–10.
- Confidence — how much evidence backs the Impact estimate, 1–10.
- Ease — how cheap and fast it is to ship, 1–10 (higher = easier).
Both conventions are in common use. The average keeps the result on the same 1–10 scale as the inputs and is the form Sean Ellis popularized for growth experiments; the product (1–1000) exaggerates gaps between ideas. Pick one and apply it to your whole backlog — never mix the two in one ranking.
Why ICE beats arguing about the roadmap
The real value of ICE is not the number, it is the conversation. Putting Impact, Confidence and Ease on the table separately stops the loudest voice from winning and forces the team to say why an idea deserves a sprint. The score is just the visible output of that shared judgment.
ICE’s weakness is also its strength: it is subjective. Two people can score the same idea differently, which is fine for a fast internal ranking but dangerous if you treat the number as a forecast. Use ICE to sort a backlog, not to promise a result — and re-score after each experiment teaches you something new about Confidence.
When ideas vary a lot in cost or audience size, ICE flatters cheap small bets and an expensive idea that would actually move the business can look the same as a quick tweak. That is the signal to switch to RICE, which divides by effort and adds reach. Many teams start with ICE and graduate to RICE as the stakes rise.
ICE bands we use as a rule of thumb
There is no official ICE cutoff — the scale is relative to your own backlog. These bands are an RGM rule of thumb for turning the average into action, not a published standard.
| ICE average | Verdict | What to do |
|---|---|---|
| 8.0 – 10 | Do it now | Front of the queue; ship this sprint |
| 6.0 – 7.9 | Schedule soon | Strong bet worth a sprint |
| 4.0 – 5.9 | Park or sharpen | Fix the weak axis or set aside |
| 1.0 – 3.9 | Drop or rethink | Low impact or confidence; not worth it |
What practitioners say about ICE
Score every idea on Impact, Confidence and Ease, rank them, and run the top of the list. The framework is meant to be fast and repeatable, not perfectly precise.
Prioritization frameworks earn their keep by making trade-offs explicit; the number matters less than the conversation it forces.