Intercom: A number should start a conversation.
Reach, impact, confidence and effort make assumptions visible. Move one input and watch the preferred option change.
See the idea in one picture ↓- the RICE score
- R × I × C ÷ E
- estimated per idea
- 4 inputs
- can still override
- Judgement
Figures as reported by the company.
Independent educational analysis. Scenarios, people, numbers and illustrations on this page are original PMcademy practice material, not company research results or an endorsement.
Different roles see different parts of the burden.
“Everyone has evidence. Nobody has the same evidence.”
- A shared way to compare
- Room for strategic bets
- Precise-looking guesses
- Dependencies a score ignores
Read the analysis
A support lead notices repeated tickets, an engineer sees implementation risk, and a salesperson remembers a lost deal. Give each person a way to contribute evidence without pretending the observations are interchangeable. Ask which population the reach estimate describes, which outcome impact changes, and what effort includes. Shared definitions matter more than an attractive spreadsheet.
A confident pitch can outrun weak evidence.
The loudest pitch wins
Assumptions are visible and comparable
Read the analysis
Imagine two roadmap proposals. One has a vivid customer story; the other affects a larger group but sounds less exciting. A shared scoring method can expose the assumptions behind the pitches. It becomes less useful if the team treats a precise-looking number as an objective truth or adjusts inputs until a favourite proposal wins.
What RICE does, and what it leaves to you
- 1Fill in four inputs
Estimate reach, impact, confidence and effort for each idea.
- 2Compare the scores
The formula turns those guesses into one comparable number.
- 3Apply judgement
A dependency can still move a lower score to the front.
Animated illustration. Names, numbers and people are fictional teaching material.
Turn the uncertain dial.
Illustrative data and simplified behaviour. This model does not represent a live company system or measured company results.
Read the analysis
Use the confidence control below. Option A reaches 100 people, has impact 2 and takes two person-months. Option B reaches 160, has impact 1, confidence 80 percent and the same effort. Both use the same reach period. As A’s confidence changes, the ordering can reverse. These are fictional inputs used to expose sensitivity, not recommended scores for a real roadmap.
Investigate the input that changes the decision.
- What Intercom says
Intercom defines RICE and allows dependencies to justify work outside score order.
- What it cannot prove
That higher-scoring ideas deliver more value in practice.
- What you would measure
Estimates against actual results
How often order is overridden, and why
Confidence scores over time
Read the analysis
If a small confidence change reverses the order, a short discovery task may be valuable. If the ranking stays stable across plausible inputs, another week refining decimals may not help. Ask the owners of the estimates for source links and uncertainty ranges. Revisit effort after technical exploration and include coordination, rollout and support where those costs matter.
Measurable work can crowd out necessary work.
- Visible assumptions
- Fairer comparisons
- Faster debates
- False precision
- Enablers undervalued
- Scores gamed to fit
Read the analysis
A scoring framework can understate enabling infrastructure, mandatory fixes, or a dependency that unlocks another project. Keep these reasons explicit rather than distorting reach or effort to make them fit. A separate strategic rationale is easier to review than an inflated score. Also record the opportunity cost: which plausible work waits if you choose this proposal?
Would you build, investigate, or override?
Your fictional team has one week before planning. You can accept the current ranking, gather evidence for the most sensitive assumption, or prioritise a dependency with an explicit rationale. Choose the action that makes the decision more defensible. Explain what evidence would change your mind and avoid using the framework as a way to end a legitimate disagreement.
Choose a next move
Build a scorecard with an audit trail.
Your artifact
Build a three-option RICE scorecard with ranges, sources, dependencies and a written recommendation.
Critique my project artifact using only the evidence I provide. Build a three-option RICE scorecard with ranges, sources, dependencies and a written recommendation. Separate observed facts, assumptions, proposed tests and missing evidence. Identify a failure mode and a cheap check. Do not invent company results or user research. Ask for my artifact before starting.
Read the analysis
Compare three project opportunities using consistent reach periods and effort units. Attach a source, confidence rationale and plausible range to each uncertain input. Add dependencies and strategic exceptions in a separate column. Finish with a decision statement and a review date so the scorecard can evolve as you learn.