Prioritise Thirty Requests in One Pass

When to use it

The user has a list of things people want done (feature requests, internal projects, customer asks, improvement ideas, a quarter's candidate work) and more of them than the team can do. They paste the list and get it ranked, with the scores visible, a line showing what fits, and a short reply for every item that does not.

Also for re-ranking when something new arrives: score the new item on the same scales and show where it lands and what it displaces.

Not for: planning the user's own week, scheduling work that is already chosen, or estimating a single project in detail. Items in the list are scored against each other, not against a budget.

What it asks you

Ask in one message, numbered. Work from what comes back.

1. Paste the list, one item per line, with who asked for each one if you know.
2. What is the goal this work serves this period, in one sentence? Impact is scored against this goal and nothing else.
3. Who are the people reach is counted in: customers, users, staff, clients? Roughly how many in total?
4. How much capacity is there for this period: people and weeks, or hours? This draws the line.
5. For each item, anything you know about size: a day, a week, a month, or a team for a quarter. Rough is fine; say "unknown" where it is.
6. Is anything on the list non-negotiable: a legal date, a contract, a fix to something broken? Those are placed, not scored.
7. Who will see the ranking: the team, the requesters, a leadership group? This sets how the not now lines are worded.

A vague answer to 2 gets one follow-up: "If only one thing on this list got done this period, what would make it the right one?" If the user still has no goal, the skill scores impact against "revenue or cost this period" and says so at the top, because a ranking without a stated goal is a ranking against a hidden one.

Never invent a reach number, a size, a requester or a deadline. "Unknown" is scored by the rule in Standards and marked so.

What comes back

Plain text.

PRIORITY RANKING: period, goal in one sentence, population and size, capacity NON-NEGOTIABLES (placed, with effort): ... Capacity left for scored items: ...

RANKED (table rows) Rank. Item (requester). Reach R, Impact I, Confidence C, Effort E. Score = R x I x C / E = value. Size. Running total. In or below the line. (one row per item, the line drawn as a row that says CAPACITY LINE with the total at that point)

RATIONALE (only where a score would surprise): item, which score, one or two lines.

NOT NOW (one line per item below the line, addressed to the requester, with the look-again date)

Swaps worth considering: at most two, each with the arithmetic (what comes out, what goes in, capacity after). Next: one line naming the single most useful next step.

For a new item arriving later: its four scores and arithmetic, where it lands, what it displaces, and the not now line for the displaced item.

• 00_START_HERE.pdf: the two-minute set-up for each AI.
• 1 - Skill: prioritise-thirty-requests.zip, one upload to Claude, ChatGPT, Gemini, Grok or Microsoft 365 Copilot. It switches on by itself when this job comes up.
• 2 - Text file for any AI: the same Move as one .txt. Attach it to any chat, on any plan, and type "Follow this file".

It drafts, you read and send. It never contacts anyone for you and never invents a fact, a number or proof.

A Move is a single play. The whole function in a box is an agent pack.