Example data. Fieldnote is a fictional company used to fill all 61 boards, so the examples tell one consistent story.
Scale-up, Karim, Tom and the data analyst score the four finalists from the impact/effort matrix before presenting the quarter to Élise.
What the team took away
Offline mode wins on reach, the invoicing integration wins on retention: the team commits to both, in that order, and parks the customer portal for a year.
How to run it
Preparation
Invite the PM, the tech lead and someone from data. Data is there to challenge Reach numbers, not to watch.
Pre-fill the table with the initiatives and a Reach figure sourced from analytics for each one. Reach guessed live is the number one cause of bad scores.
Write the scales on the board before starting: Impact 3 / 2 / 1 / 0.5 / 0.25, Confidence 100 / 80 / 50 %, Effort in person-months.
Agenda
10 min
Align on the scales Read the scales aloud and pick one initiative everyone knows as the reference: 'this one is Impact 2, Effort 1'. Every other score is relative to it.
20 min
Reach Go down the list. For each initiative, data confirms or corrects the number of users touched per quarter. Write the source next to the figure.
40 min
Impact and Confidence PM proposes, the table challenges. Confidence drops to 50 % whenever nobody can cite evidence. Do not let anyone put 100 % on an idea nobody tested.
30 min
Effort Tech lead estimates in person-months including design and QA. Round to the nearest 0.5. If two initiatives get the same effort, ask which one they would rather build first and why.
15 min
Compute and read Fill the Score column. Read the top 5 aloud and ask: does this order surprise anyone? If yes, check the inputs, not the formula.
5 min
Decide Freeze the top 5 for the quarter. Note which scores relied on a Confidence below 80 % and plan the experiment that would raise it.
Pitfalls
Do not tune the scores until the order matches what the loudest person wanted. If the table needs tuning, the scales were unclear: fix the scales.
Keep Reach as a count of users, never a percentage. Percentages hide the fact that 40 % of a small segment is nothing.
Score in one session and stop. RICE is a conversation starter, not a spreadsheet you polish for a week.
Afterwards
The top 5 feed the roadmap or the Now column of a Now / Next / Later. Anything scored with low Confidence goes to an experiment board before it gets built.