Best for
- Backlogs ordered by whoever asked loudest or latest
- The quarterly planning that needs an order defensible to stakeholders
- Saying no with reasons instead of avoidance
What you give it
- The candidate list, your strategy context, and honesty when the scoring hurts a favourite
What you get back
- Every candidate scored on the fitted model: value evidence, effort reality, risk and confidence — assumptions written down, not buried in the numbers
- The order with its reasoning per item — and the cut line: what falls below it, said explicitly, with the why
- The disagreement surface: where scores and gut diverge, surfaced for the conversation that actually decides
How it works
- Fits the model to the strategy first: this quarter's weights (retention, expansion, debt, risk) made explicit — a generic formula smuggles in whoever wrote it.
- Scores with evidence and confidence separately: the value claim, its evidence, and how sure — low-confidence high-value items become validation spikes, not leaps.
- Estimates effort against your delivery history, not optimism — and in ranges, because false precision breaks the whole ordering.
- Surfaces the divergences deliberately: where the scores fight the gut or the politics, the pass makes the conflict visible and arguable instead of silently resolving it either way.
Example
You: Prioritise these 31 backlog items for the quarter; currently the order is 'CEO asked' then 'biggest customer asked'.
Result: The pass: a model fitted to the stated strategy (retention-weighted this quarter — expansion items score against it honestly), each item scored with its evidence and confidence noted (the retention-dashboard rebuild scored high on value but LOW confidence — flagged for a one-week validation spike instead of a ten-week build), effort estimated against history not hope. The uncomfortable findings handled on paper: the CEO's pet item scored 19th (the conversation happened with the scoring visible — it moved to a cheap partial version), two sacred-cow maintenance items scored above every new feature (funded first), and the cut line after item 12 said out loud with reasons. The quarter shipped the top 11; the defence deck was just the scoring sheet.
Limits — please read
- Scoring structures judgement; it does not replace it — the final call on genuine ties and strategic bets stays human, on the record.
- Garbage estimates in, garbage order out: where value evidence is missing, the honest score is 'unknown — validate first'.
- The cut line needs saying by someone with authority; the pass writes the reasons that make saying it survivable.