RICE (Intercom)
Reach (how many people in a time period) × Impact (1, 2, 3 — minimal/medium/high) × Confidence (% — "I have data" vs. "I have a hunch") ÷ Effort (person-months). Higher = better. Best for product features where reach + impact are estimable.
WSJF (SAFe)
Weighted Shortest Job First: Cost of Delay ÷ Job Size. Cost of delay = user/business value + time criticality + risk reduction / opportunity enablement. Best for portfolio-level prioritization where the question is "which big bet first" not "which feature."
MoSCoW
Must have (failure if we miss), Should have (important, not critical), Could have (nice to have), Won't have (this time). The won't column is the load-bearing one — it forces explicit deprioritization. Most useful for fixed-deadline projects where you're negotiating scope.
↳ all three are imperfect