framework · prioritization
Weighted scoring prioritization
Best for: Decisions where the dimensions of value are genuinely different in kind (strategic bets, platform choices, vendor selection, internationalization order) and RICE's four fixed variables cannot sort the candidates correctly.
Weighted scoring is a prioritization method where you define the criteria that matter for a specific decision, assign each a weight (all weights must sum to 100%), score each option on each criterion, multiply score by weight, and sum to a final number. The result is a ranked list. The reason to reach for this tool is not that it is more rigorous than RICE: it is that RICE has a fixed schema (Reach, Impact, Confidence, Effort) built around user-facing impact-per-effort, which means RICE cannot hold strategic alignment, regulatory risk, vendor lock-in, or brand fit. When those criteria are the actual decision drivers, RICE gives you a confident wrong answer.
When RICE fails and weighted scoring is the right call
RICE forces you to express every priority as a ratio of user impact to engineering cost. In 2026, that ratio has a structural problem: the Effort denominator has collapsed. A feature that took a month of development in 2023 may take a few days with AI-assisted development. When Effort no longer sorts candidates meaningfully, the output of a RICE calculation is noise wearing the costume of data.
More importantly, RICE has no slot for dimensions that now do meaningful sorting work: whether a bet deepens a competitive moat or just adds surface area; whether there is a paying market for the outcome; whether users will adopt the result or merely tolerate it. Those are viability and lovability questions, not effort questions.
Reach for weighted scoring when:
- Multiple stakeholders disagree about priorities because they are weighting different criteria, not because they disagree about scores
- The candidates being compared differ in kind, not just in magnitude (vendor A vs. vendor B, market X vs. market Y)
- Strategic alignment, regulatory exposure, or market size is a primary decision driver
- You are making a platform or infrastructure bet where reversibility cost matters more than current user volume
The process
1. Name the decision precisely. “What should we build next?” is not a decision. “Which of these three AI features should anchor Q3 for our growth segment?” is.
2. Define criteria together. Run a short session where each stakeholder proposes criteria and the group negotiates. This step is not administrative: the facilitation move of negotiating criteria as a group turns the final artifact from “the PM’s spreadsheet” into a shared contract. When a VP of Sales later argues for a different outcome, they are arguing against weights they helped set.
3. Assign weights, summing to 100%. If you cannot justify a specific allocation from strategy or data, equal weights are the recommended default. Equal weights are less gameable than a distribution someone with authority invented in the moment. Common error: giving the heaviest weight to the criterion the most senior person cares about.
4. Score each option on each criterion, 1 to 10. Anchor the scale. “10 means fully aligned with the 18-month roadmap as written” is a score. “10 means great” is not.
5. Multiply score by weight, sum across criteria. Rank by total score.
6. Interrogate the result before accepting it. Does the top-ranked option feel wrong? That is information. Either the weights are off, or you have a criteria definition problem.
A worked example
Decision: which of two integrations to ship first for a B2B workflow tool.
| Criterion | Weight | Zapier integration | Streak CRM |
|---|---|---|---|
| Cumulative MRR from requestors | 40% | 8 | 4 |
| Implementation ease | 25% | 9 | 3 |
| Strategic alignment (18-mo roadmap) | 25% | 7 | 5 |
| Risk (security, compliance exposure) | 10% | 9 | 6 |
| Weighted total | 7.95 | 3.95 |
Zapier scores 7.95. Streak CRM scores 3.95. The gap is large enough that the ranking holds even if you redistribute 10 points of weight. That stability test matters: if re-weighting plausibly changes the winner, the decision is not yet resolved, and the session should continue on criteria rather than scores.
This is Savio’s real worked example (with rounded weights for legibility), which is useful precisely because it is unglamorous: the winning candidate was obvious from MRR data, and the model confirmed it. That is the correct use of weighted scoring: making the reasoning visible, not discovering a surprise.
The failure modes
Garbage criteria in, confident garbage out. If you define criteria that sound meaningful but cannot be scored independently (“customer delight,” “innovation potential”), you get laundered intuition with decimal points. Spend more time on criteria definition than on scoring.
Weight gaming by authority. If the heaviest weight goes to the criterion the most powerful person in the room cares about, the model is a rationalization device. The counter-move: set weights before scoring any specific option, and document the rationale. Changes after options are visible require explicit justification.
Missing the customer voice. Strategic alignment and revenue potential are easy to overweight in stakeholder sessions. Name a “user lovability signal” criterion explicitly: would users actually adopt this, or just tolerate it? That question forces the group to distinguish reach from adoption from delight.
Incommensurable criteria forced into a common scale. Regulatory compliance is not commensurable with user delight on a 1-to-10 scale. When one criterion is a binary gate (pass/fail, legal requirement, hard dependency), pull it out of the weighted model entirely and treat it as a filter before scoring.
Use it, do not recite it
The risk of weighted scoring in an interview is the opposite of RICE’s risk. RICE answers sound mechanical because the formula is familiar. Weighted scoring answers sound hollow when the candidate names the steps without naming the facilitation move. The framework is useful only if you have actually run a criteria-negotiation session and seen what it does to stakeholder dynamics.
A strong answer names the trigger condition that made weighted scoring the right choice, describes setting up criteria collaboratively as a deliberate design decision, walks through a real or realistic example with specific criteria and weights, names a failure mode they have actually navigated, and closes by explaining when they would choose RICE or Kano instead.
A weak answer lists the steps mechanically and says the framework is “more flexible than RICE because you can pick your own criteria,” uses generic criteria (customer impact, technical feasibility, revenue potential) without explaining the selection, and treats the output as the answer rather than as a prompt for the harder conversation.
The model does not make the decision. It structures the argument so the decision can be made on the right terms.
Related: RICE · ICE · Kano model
Related
- RICE prioritization framework prioritization
- ICE scoring framework prioritization
- Kano model (and how to use it in a 2026 PM interview) prioritization