Prioritization

Every backlog is infinite. Capacity is not. Prioritization is the craft of deciding what not to build so that what you do build has a chance to matter.

A scoring framework will not save a team without strategy. It will force honest conversation about reach, impact, confidence and cost — and surface when "must-have" work is actually a low-confidence guess dressed as urgency.

Start from product strategy: a guiding policy that already killed whole categories of work. Then rank what remains. Without strategy, RICE ranks a random pile with false precision.

A picture of the arithmetic: onboarding that reaches 8,000 new users can score 12,800 on RICE; custom roles for 400 enterprise users can score 360 even when sales is loud. The score is not a verdict — it might miss $2M of pipeline or a p95 that is killing retention — but it forces the unit of value onto the table. Then reserve capacity so reliability and debt are not staffed only after incidents. A short decision log next to the ranked list stops the same Monday argument from starting over.

When asked how you prioritize, walk through factors (user value, confidence, cost, strategic fit, risk), give a tiny RICE example, then say what you would not trust a score for (compliance, outages, ethics). Mention opportunity cost in one sentence.

Concepts in this section

Step	Concept	What it answers
1	Opportunity cost	What does saying yes actually cost?
2	RICE scoring	How do you compare optional bets on one scale?
3	ICE, Kano and WSJF	What do scores miss: basics, delay, similar-reach experiments?
4	Capacity allocation	How much of the quarter is features vs debt vs reliability?
5	Cheaper versions	How do engineers change the ranking without shouting?

Pick one framework and stick with it for a quarter so scores stay comparable. Publish the ranked list. When leadership overrides, write the reason next to the item. Pair with roadmaps: Now should be mostly top-scored or explicitly overridden items; park the rest in an icebox with a revisit date.

Common mistakes

Everything is P0. Nothing finishes.
Precision theater and gamed Confidence.
RICE on a broken checkout. Basics and SEVs are not optional bets.
No decision log. The same debate repeats every Monday.

Key takeaways

Prioritization is choosing what not to do; opportunity cost is the real price.
RICE = Reach × Impact × Confidence ÷ Effort — relative scores and honest units.
ICE is faster when reach is similar; Kano protects basics; WSJF / cost of delay favors short, high-urgency jobs.
Allocate capacity across growth, commitments, reliability and debt.
Treat frameworks as structured debate; document overrides. Engineers win with honest effort and cheaper slices.