Skip to content
PMcademy
Drill75 minutesUnit 6 of 8 · Lesson 3 of 3

Which one is worth a quarter?

Unit 6 of 8 · Lesson 3 of 3Index

Five real problems, one team, twelve weeks. Maya wants all five, and she said so in front of everyone.

Confidence is the honest column

  1. All five, in twelve weeks

    01 All five, in twelve weeks

    The empty bank sync, the AI assistant for everyone, a premium tier, calmer notifications, rent reminders. Maya wants every one.

  2. Score them the same way

    02 Score them the same way

    How many people, how much it helps, how sure you are, how much work. Same scale, same sheet.

  3. The column nobody fills in honestly

    03 The column nobody fills in honestly

    The assistant looks unbeatable until you ask what you actually know. Nobody has tested it, so its confidence drops to what you can prove.

  4. Watch the order move

    04 Watch the order move

    Set every confidence to the evidence and the ranking changes. The framework did not decide. It showed you where to argue.

  5. The list nobody wants to read

    05 The list nobody wants to read

    What you are not doing this quarter, with one line on why. It is the most useful part of the page.

Try the concept

How much confidence changes the choice?

Interactive teaching example · All numbers and tickets are fictional.

AB0100A: 50 · B: 64 · Option B ranks higher

A = 100 reach × 2 impact × 0.5 confidence ÷ 2 person-months. B = 80 × 1 × 0.8 ÷ 1 = 64. A needs confidence above 64% to lead. Evidence must justify that change.

How Intercom handled it

The point of the score is not the score. It is making a team say out loud how much it actually knows.

Intercom
The situation

Intercom's product team kept arguing about priorities, and every favourite idea arrived with a confident champion.

What they did

Sean McBride, a product manager there, wrote up the scoring they settled on: reach, impact, confidence and effort. Confidence was in it on purpose, to rein in ideas that were exciting but unproven.

Public commentary. No affiliation.

Doing it with AI, and where it breaks

The move

Build the scoring sheet, then ask which scores are not supported by the evidence you described.

The trap

It will accept your numbers and do the arithmetic, which flatters you. Make it attack the inputs, not compute the output.

Prompt worth stealing

Here are five opportunities with reach, impact, confidence and effort, and the evidence behind each one [paste]. Do not compute scores. For each confidence value, say whether the evidence supports it and what number it does support. Then tell me which items would change rank.

Build it

The artefact

A scored shortlist where every confidence is backed by evidence, and a not doing list.

Google Sheets

0/4

The worked solution

Stuck, or want to compare?

There is a worked solution. Try the build first. Reading it before you have attempted anything is the fastest way to learn nothing.

Checkpoint

If you did the build, these take two minutes. If you cannot answer one of them, that is the part to go back to.

Go deeper

Moving on marks this lesson complete. Finish the build first, it is the part that counts.

Previous lessonNumbers without a data team