Every prioritization framework exists to turn an argument into a number you can defend. RICE, MoSCoW, WSJF, Kano, a weighted scoring model — they're different lenses on the same question: what should we do next, and how sure are we? This page gives you each one with the actual formula and a worked example, then a straight table of which to reach for when. And then the part the templates skip — a score is only true until one of its inputs changes, and in a spreadsheet nothing tells you when that happens.
Prioritization is just making the trade-off explicit instead of implicit. You have more good ideas than capacity; a framework forces you to say why this one beats that one in terms you can repeat to a skeptical stakeholder. The number isn't the point — the agreed inputs behind the number are. A RICE score of 42 means nothing; "reach 8k, impact 2, confidence 80%, effort 3 person-months" means something, and it's arguable in the right way.
Which is exactly why prioritization decays. Those inputs are assumptions about a moving world. Reach was estimated before the segment shifted; effort was a guess before the spike; confidence was high before the experiment came back flat. The ranking you defended in January is quietly wrong by March — and the spreadsheet looks just as confident as the day you filled it in. A framework gives you a good decision; nothing in a spreadsheet keeps it good.
Each with its formula, a one-line worked example, and the situation it's built for. None is "best" — they trade rigor for speed differently.
Score = (Reach × Impact × Confidence) ÷ Effort. E.g. (8,000 users × 2 × 0.8) ÷ 3 = 4,267. Best when you have rough quantitative inputs and want a comparable score across very different features. The confidence term is the honest part — it docks ideas you can't back up.
Sort every item into Must-have, Should-have, Could-have, Won't-have (this time). No math — it's a fast alignment tool, ideal for scoping a release with stakeholders in a room. Its discipline is the "Won't" bucket: naming what you're deliberately not doing prevents silent scope creep.
Weighted Shortest Job First = Cost of Delay ÷ Job Size, from SAFe/agile. Surfaces the small, time-sensitive things that a pure impact score buries. Best when sequencing a backlog where timing matters — a fix worth a lot now and little later scores high.
Classifies features by how they affect satisfaction: basics (absence hurts, presence is invisible), performance (more is better), delighters (unexpected upside). Best for balancing a roadmap so you don't over-invest in delight while a basic expectation is missing.
A decision matrix: pick criteria (impact, effort, strategic fit, risk), weight them, score each option, sum. The most flexible — you define what matters. Best for a one-off, high-stakes choice between a few options where RICE's fixed inputs don't fit. (This is the "weighted decision matrix".)
Match the tool to the decision, not the other way around. A quick guide.
| Your situation | Reach for | Why |
|---|---|---|
| Comparing many features with rough data | RICE | One comparable score, confidence-adjusted |
| Scoping a release with stakeholders | MoSCoW | Fast alignment, forces a "Won't" list |
| Sequencing where timing matters | WSJF | Surfaces small, urgent, cheap wins |
| Balancing basics vs delight | Kano | Classifies by satisfaction effect |
| One high-stakes choice, few options | Weighted matrix | You define the criteria and weights |
A prioritization score isn't a fact — it's a decision built from inputs. The framework is only as live as those inputs, and a spreadsheet has no idea when one of them moved. This is the narrow thing Draftlize does.
Reach, impact, confidence, effort — each is an assumption that lives somewhere. In Draftlize they're cards: a reach estimate cites the analysis behind it, a confidence level cites the experiment that set it. The score links to its inputs instead of hard-coding their values in a cell nobody revisits.
The experiment comes back flat, so confidence drops. The card that held the old confidence changes, and every score that depended on it auto-flags stale — so a re-prioritization is triggered by the world moving, not by someone happening to re-open the sheet. The ranking stays honest because the substrate, not your memory, tracks what it rests on.
Ask Claude Code or Cursor to re-rank the backlog and it reads the current inputs over MCP — the live reach, the updated confidence — not the values you typed last quarter. The framework becomes something that re-runs against reality instead of a snapshot you defend long after it stopped being true.
A prioritization framework gives you a good decision. Nothing in a spreadsheet keeps it good when the numbers underneath it move.Keep your framework. Track the inputs it rests on so the ranking can't quietly rot.
MoSCoW. It needs no math and no data — you sort items into Must, Should, Could, and Won't-have. It is the fastest way to get a room aligned on the scope of a release, and you can graduate to RICE once you have rough quantitative inputs.
RICE optimizes for overall value per unit of effort and is confidence-adjusted, so it is good for comparing many features broadly. WSJF (Cost of Delay ÷ Job Size) optimizes for sequencing when timing matters, so it surfaces small, urgent, cheap wins that RICE would bury. Use RICE to choose, WSJF to order.
You do not — they are estimates, and pretending otherwise is the trap. The value is in writing the assumption behind each number so it can be challenged and revisited. A confidence of 80% should cite why; when the why changes, the score should change with it.
When an input changes, not on a fixed calendar. A monthly re-score is a workaround for not knowing which assumptions moved. If your inputs are tracked, re-prioritization is triggered by the thing that actually invalidated the old ranking.
Score with whatever framework you like. Put the inputs in Draftlize as cards, and let a changed assumption flag every ranking that depended on it — so your priorities re-rank when the world moves, not when you remember to.
Start free with $5