Product Management prompt of the day

Opportunity sizing: estimate reach and impact with every assumption in the open

Someone says a feature will move the business and the number is a guess. Build a sizing model where each assumption is visible and can be challenged.

You are a product strategist sizing a product opportunity. I will describe the feature, the users it serves, and whatever data I have.

Produce:

A) SIZING TREE — the estimate broken into multiplicative steps (eligible users x share who notice x share who adopt x effect per adopter), each step with its value, source, and a confidence label (measured / benchmark / guess).

B) RANGE — low, base, and high cases built by changing the weakest assumptions, with the arithmetic shown.

C) SENSITIVITY — the one or two assumptions that move the result the most, and how to measure each cheaply.

D) COMPARISON NOTE — how this estimate would need to change to beat the next item on the roadmap, so the team knows what evidence would matter.

Inputs: [FEATURE DESCRIPTION] · [TARGET USERS + COUNTS IF KNOWN] · [ANY USAGE OR CONVERSION DATA] · [THE NEXT ROADMAP ITEM IT COMPETES WITH]

Rules: Do not invent user counts, conversion rates, or benchmarks — use my data, mark every other value as a guess, and show the math for each step. Verify the measured inputs against your analytics before the estimate is shared. Keep confidential metrics and customer data out of consumer AI tools beyond what I provided.

Why it works — Opportunity sizes are often a single number with hidden assumptions, which makes them impossible to challenge and easy to inflate. A multiplicative tree with a source and confidence label on each step shows exactly where the estimate is weak, and the sensitivity section turns the weakest link into a cheap measurement. The range replaces false precision with an honest spread.