1. What are you choosing?
Direction, problem, bet, experiment, release scope, customer-response evidence, or sequence are different decision objects.
Product prioritization workflow
Product prioritization: choose the decision before the framework
Prioritization is not one decision. First identify what is competing, whether the candidates are actually comparable, and what output you need. Then use the method or tool that fits that job.
Framework choice comes after decision definition. If the candidate set is broken, fixing the set is part of prioritization.
2. Are they comparable?
Check decision level, outcome, horizon, evidence, and hard constraints before a score enters the conversation.
3. What should exist afterward?
A strategy choice, credible opportunity, explicit bet decision, scoped release, Kano evidence, or roadmap are different outputs.
Choose the right prioritization approach
What are you actually trying to prioritize?
Start at the decision level, not at the framework name. Each route below needs different inputs and leads to a different result.
Direction
The strategic direction is still unresolved
You are choosing between markets, customer segments, outcomes, growth motions, or other strategic directions-not between comparable delivery candidates.
- Ready when
- A target customer/outcome, a strategic choice, explicit boundaries, and non-goals.
- Expected output
- A direction that constrains what should enter the prioritization set.
- Watch out for
- Do not use a feature score to manufacture strategy after the fact.
Problem / opportunity
The underlying problem is not credible enough yet
You have many requests or solution ideas, but the customer problem, opportunity, evidence, or assumptions underneath them are still weak.
- Ready when
- A clear problem/opportunity, supporting evidence, assumptions, and the unknowns that still matter.
- Expected output
- A decision-ready problem or opportunity-not a more precise ranking of weak assumptions.
- Watch out for
- Do not rank solutions precisely when the problem case itself is unresolved.
Bets / initiatives
Comparable product investments are competing for capacity
The candidates are plausible investments at roughly the same decision level and they compete for the same scarce capacity or outcome.
- Ready when
- Decision context, evidence, uncertainty, strategy, constraints, dependencies, and opportunity cost.
- Expected output
- An explicit Select / Learn / Sequence / Defer / Reject-style decision with rationale and a revisit trigger.
- Watch out for
- Do not treat the top numerical score as an automatic commitment.
Relative order
You need a quantitative relative comparison
The candidates are genuinely comparable and a shared set of inputs can make the trade-off more legible without pretending the estimates are facts.
- Ready when
- A shared outcome, horizon, input definitions, and evidence strong enough for the chosen method.
- Expected output
- A relative order to review and challenge-not objective truth.
- Watch out for
- Do not score unlike objects, incompatible horizons, or unsupported precision.
Release scope
The real job is negotiating what fits a bounded release
The outcome or release is already chosen, capacity is constrained, and the argument is about what is truly required now versus what can move out.
- Ready when
- A real release/outcome boundary and a defensible definition of what failure makes something a Must.
- Expected output
- Must / Should / Could / Won't-this-cycle scope with explicit trade-offs.
- Watch out for
- Do not use MoSCoW as a generic portfolio ranking formula.
Customer response
The question is how capabilities affect customer expectations
You need to understand whether capability presence or absence behaves like a baseline expectation, performance driver, delighter, or another Kano response pattern.
- Ready when
- Paired customer-response evidence that can support Kano classification.
- Expected output
- Customer-response evidence that can inform a later investment decision.
- Watch out for
- Kano does not decide strategy, capacity, effort, opportunity cost, or roadmap sequence by itself.
Sequence
The bets are selected; now you need to communicate sequence
You already know which investments earned capacity. The remaining job is to communicate horizons, rationale, evidence, uncertainty, and dependencies.
- Ready when
- Selected bets and the reasons they earned capacity.
- Expected output
- A roadmap that communicates the chosen sequence rather than re-running prioritization inside the roadmap.
- Watch out for
- Do not populate a roadmap first and hope the roadmap itself resolves the priority decision.
Before you compare
Can I compare these candidates yet?
A framework cannot fix a malformed candidate set. If several checks fail, clarify the decision before scoring it.
Same scarce decision
The candidates should compete for the same capacity, decision, or allocation. A mandatory migration and an optional feature may coexist in planning without being the same scoring object.
Same abstraction level
Do not casually rank a strategy against a feature, a problem against its solution, or an experiment against a production rollout.
Shared outcome / context
Impact only has meaning when candidates are judged against a compatible outcome and decision context.
Compatible horizon
A two-week learning test and a multi-year strategic investment should not be flattened into one score without acknowledging the mismatch.
Evidence fits the method
If the inputs cannot support precise numbers, use coarser judgment or learn first rather than converting guesses into decimals.
Hard constraints separated
Legal, security, contractual, continuity, or mandatory migration constraints can change the feasible set instead of behaving like ordinary value factors.
Compare the methods
Choose the method that fits the trade-off
These methods are not interchangeable versions of the same score. Use the one whose inputs and output match the decision you already defined.
Use when
Comparable investments compete for scarce capacity.
Needs
Evidence, uncertainty, strategy, constraints, dependencies, opportunity cost.
Produces
A human decision, rationale, override logic, and revisit trigger.
Do not use when
Do not use it to repair unresolved strategy or a weak problem case.
Use when
Reach, Impact, Confidence, and Effort can be interpreted consistently across the set.
Needs
A shared horizon and meaningful definitions for all four inputs.
Produces
A relative numerical order to review.
Do not use when
Avoid when Reach is artificial or candidates are not comparable.
Use when
A faster, coarser relative comparison is useful and detailed Reach modeling adds little value.
Needs
Consistent Impact, Confidence, and Ease/effort interpretations.
Produces
A lightweight relative order.
Do not use when
Avoid when a quick score would disguise a missing strategic or evidence decision.
Use when
You are negotiating scope inside a bounded release, outcome, deadline, or capacity envelope.
Needs
A real scope boundary and a defensible definition of Must.
Produces
Must / Should / Could / Won't-this-cycle scope.
Do not use when
It is not a generic mathematical ranking system.
Use when
The unknown is how capability presence/absence changes customer satisfaction or expectations.
Needs
Paired Kano responses from relevant customers or segments.
Produces
Customer-response classifications and interpretable evidence.
Do not use when
It informs prioritization; it does not settle the whole investment decision.
Product decision workflow
Strategy → evidence → decide → sequence → define → learn
Prioritization sits between credible upstream choices and downstream communication. A roadmap does not replace the choice; a PRD should not define work before the bet earns capacity.
Step 1
Strategy
Choose direction, target, outcome, advantage, boundaries, and non-goals.
Clarify directionStep 2
Evidence / opportunity
Turn signals and requests into a credible problem or opportunity with assumptions visible.
Strengthen evidenceStep 3
Compare / decide
Compare decision-ready candidates under scarce capacity and record the actual trade-off.
Make the decisionStep 4
Sequence / communicate
Communicate selected bets, horizons, uncertainty, dependencies, and what is not now.
Build the roadmapStep 5
Define selected work
Only after a bet is selected, frame its outcome, boundaries, open questions, and then detailed requirements as needed.
Frame the selected betStep 6
Learn / revisit
Use evidence from execution or experiments to revisit the decision when assumptions, constraints, or outcomes materially change.
Practice the judgment
Do the right kind of work
Live decision, framework learning, practice, and interview are separate jobs
Make a live Product decision
Use real evidence, constraints, opportunity cost, uncertainty, and human judgment.
Open the workspaceLearn framework mechanics
Compare RICE, ICE, MoSCoW, Kano, Value vs Effort, and WSJF by the decision they expose.
Read the framework guidePractice judgment
Make a prioritization choice on fictional evidence before seeing guidance and feedback.
Practice in PM ExercisesPrepare for an interview
Use the prioritization interview guide for prompts, answer structure, pressure follow-ups, and evaluation logic.
Open the interview guideWhen evidence is the gap
Strengthen discovery before adding more scoring
If the candidate opportunities are weak, Product Discovery is the deeper learning route. Use it to strengthen how problems, evidence, opportunities, assumptions, and learning connect before committing capacity.
Explore Product DiscoveryWhen the mental model is the gap
Connect outcomes, evidence, trade-offs, strategy, and roadmapping
Product Management Foundations is the broader route when the issue is not one framework but how Product decisions connect across the workflow.
Explore Product Management FoundationsThree rules worth carrying into every prioritization decision
- Framework choice comes after decision definition. A familiar formula is not a substitute for naming what is actually being decided.
- If candidates are not comparable, fix the set. Clarifying strategy, evidence, decision level, or horizon is part of the work.
- The output is not always a ranked backlog. A valid outcome can be Learn, scope a release, classify customer response, reject, defer, or communicate sequence.