Imagine having Henry Dreyfuss rank your roadmap.
He will tell you what to cut.
Dreyfuss refused to design for an imaginary user. He measured real people first and built the reference charts the whole discipline still uses. He does not ask which feature is most requested. He asks which one answers something real.
That the loudest demand is not the highest risk.
Your demand list is sorted by volume, because volume is the one thing a request tracker can count. Volume is a measure of how many people asked. It is not a measure of what happens if you do not build it.
The item hundreds of accounts asked for is real demand. The item two accounts named in signed agreements is an obligation with a date. Only the first of those facts is in your roadmap tool, so the ranking is systematically biased toward the loud and away from the binding.
The rarest and most useful thing this seat does is argue for removal. Every prioritization framework produces an ordered list of things to build; almost none of them will name the item you should drop, because nobody's incentive is served by saying so.
Somebody joins it by hand, once a cycle.
The roadmap tool holds the requests. The commercial systems hold the revenue attached to each account. The contracts hold which of those requests are commitments. Engineering holds what each one actually costs.
Someone joins all four before each planning session, in a spreadsheet, and the join is accurate for exactly as long as the session lasts.
Its standing register, derived continuously.
Every finding arrives with the records it came from: the invoice, the ticket, the message. You are never asked to trust a conclusion on its own.
SEE IT ON YOUR OPERATION
What would this seat have to catch?
Tell us the thing on your roadmap you suspect should be cut, and what evidence would settle it.
Request a walkthrough