Decide what not to build this quarter, not just what to prioritize
A roadmap is really a cutting exercise wearing a prioritization costume.
- Use it for
- PMs and founders staring at a backlog longer than any team could ship.
Most prioritization exercises produce a ranked list, which quietly avoids the actual hard question: what are you telling stakeholders no to. A ranked backlog where everything is "priority 4" or lower still implies it'll get done eventually, and that fiction is what lets a roadmap balloon past what any team can actually ship. The useful output isn't a ranking, it's an explicit no list with reasons attached.
This prompt forces the cut list into existence, with reasoning, instead of letting ranking stand in for a real decision.
When not to use this
If your backlog already fits comfortably in a quarter with room to spare, this is solving a problem you don't have. This is for the actually common case — more good ideas than capacity — where the ranking itself is the avoidance.
Did this work?
You can name specific things you're explicitly NOT doing this quarter, and the reasoning behind each cut — not just a ranked list of what you're doing.
Tested on claude-opus-5. Evidence status is draft; it moves to battle-tested only on recorded runs, never by hand.