---
name: board_paper_drafting
description: Draft board papers that get read and decided: decision framing, a one-page executive lead, honest option analysis with tradeoffs, risk and finance sections, and a draft resolution ready to pass. Use when writing, rewriting, or reviewing a board paper, board report, decision paper, or paper for a committee, trustees, or directors.
---

# Board Paper Drafting

## Purpose

Help a director who was not in the day-to-day work decide well, fast. A board paper exists to get a decision, record advice, or give assurance. It does not exist to prove how much work management did.

Directors read many papers before one meeting, often days apart from the discussion. The paper must stand alone, put the ask first, show its workings on options and risks, and hide nothing material in an appendix or a vague adjective.

Use this skill to draft a new paper, rewrite a weak draft, or review a paper before it goes to the chair or company secretary.

## Workflow

### 1. Pin down the paper type and the ask

Before writing, get these facts. Ask for any that are missing. Do not invent them.

- Paper type: For Decision, For Discussion, For Noting, or For Assurance
- Exact decision or outcome wanted, in one sentence
- Meeting date, agenda item, sponsor, author, and who approved the paper for submission
- Deadline: what breaks, costs more, or becomes unlawful if the board waits one cycle
- Board authority: is this within the board's reserved matters, a delegation limit, or a policy? Name the policy, strategy, or prior resolution it links to
- Money: total cost, one-off and recurring, budgeted or unbudgeted, funding source
- What the board decided or discussed on this topic before, with dates

If there is no decision, say so up front. A noting paper that buries a request for approval in paragraph nine will be sent back.

Test the framing: if the title and first paragraph were all a director read, would they know what they are being asked to do, why now, and what it costs? If not, reframe before drafting the rest.

### 2. Write the one-page executive lead first

Write this last in your head but first on the page. Limit it to one page. It must stand alone.

Include, in this order:

1. Purpose in one or two sentences
2. The ask: "The Board is asked to approve..." with the draft resolution wording
3. Why now: the trigger, deadline, or change that brings it to this meeting
4. Recommendation in one sentence, plus the two or three strongest reasons
5. Cost and funding in numbers, not words like material or significant on their own
6. Top two risks, and the risk of doing nothing
7. What happens next if approved: owner, first milestone, report-back date

Weak lead: opens with history from two years ago, reaches the ask on page three, uses size words with no numbers.
Strong lead: names the amount, the decision, and the consequence of delay in the first paragraph. See `references/worked-examples.md` for a paired example.

### 3. Build the body in decision order

Keep the main body short. Three to five pages is a workable target for most decision papers, with detail moved to appendices. Boards set their own limits, so follow the house template where one exists. Length is a discipline test: a long body usually means the options have not been sorted yet.

Use this order:

1. **Cover sheet**: title (one line), meeting date, agenda item, paper type, sponsor, author, date prepared, classification, strategic priority or policy link
2. **Draft resolution**: precise wording the board can pass without amendment, each separate decision listed on its own line
3. **Executive lead**: as in step 2
4. **Background**: only what a new director needs to decide. Prior board discussion, current position, what changed. Keep it tight, push chronology and detail to an appendix
5. **Options analysis**: see step 4
6. **Recommendation and rationale**: why this option wins on the stated criteria, and what tradeoff the board is accepting
7. **Implications**: finance, risk, legal and regulatory, people, stakeholders, strategy fit. Half a page each at most in the body, detail in appendices
8. **Implementation**: owner, timeline, resources, success measures, escalation trigger, and when the board hears back
9. **Appendices**: business case, data tables, external advice, draft contracts. Label each appendix and refer to it by number from the body. Never put a material fact only in an appendix

Head each section so a director can jump to it without losing the thread. Number paragraphs in long papers if the board's house style does that.

### 4. Do honest option analysis

Present two to four credible options. Always include:

- **Do nothing / status quo**: cost, risk, and what deteriorates if the board waits
- **The recommended option**
- **At least one real alternative** that a reasonable director could pick

Do not add a straw option that no one would choose just to make the recommendation look good. If an option was considered and ruled out early, list it in one line with the reason.

For each option give the same facts, in the same order:

- What it is, in two or three sentences
- Cost: one-off, recurring, and over what period
- Benefit, tied to a stated goal or criterion
- Main disadvantage, stated plainly
- Risk and reversibility: can it be undone, and at what cost
- Who is affected

Then compare options against criteria named before the scoring, for example cost, risk, speed, strategy fit, and legal exposure. If you weight criteria, show the weights. If two options are close, say they are close and name the judgement call that separates them.

Separate three kinds of statement and label them where it helps:
- **Fact**: confirmed data, with source and date
- **Assumption**: estimate or forecast used in the analysis, with the range if there is one
- **Judgement**: management view, owned as a view

Example of an honest line: "Assumption: take-up reaches 20 percent in year one, based on the pilot in March. At 10 percent the project breaks even in year three instead of year two."
Example of a line to cut: "The impact will be significant and synergies are expected." No number, no source, no owner.

### 5. Write risk and finance so a director can challenge them

Finance:
- Quantify every material impact. Use ranges where the number is uncertain and say why it is uncertain
- Split capital and operating cost, one-off and recurring, budgeted and unbudgeted
- State cash effect and timing, not only the accounting total
- State what is displaced: if this is funded, what does not get funded

Risk: use a short table in the body.

| Risk | If it happens | Likelihood | Mitigation | Residual level | Owner |
|---|---|---|---|---|---|
| Supplier misses go-live date | Service gap of up to 4 weeks | Medium | Contract milestone payments, fallback supplier on standby | Medium | Operations Director |

Add a line for the risk of not acting. Name uncertainty that remains after mitigation. Do not lower a risk rating just to make the recommendation easier to approve.

### 6. Draft the resolution like a minute

Good resolution wording is specific, dated, and complete enough to act on without the paper beside it.

Pattern: "That the Board approves [action] at a cost of up to [amount], funded from [source], with [named role] authorised to [specific authority], and asks for a progress report at the [month] meeting."

Avoid vague verbs in a resolution: note, consider, and endorse are for noting and discussion papers. Approve, direct, and authorise are for decisions. List each decision separately, as UK Parliament board guidance requires, so the minutes can record each one cleanly.

### 7. Run the challenge test before sending

Rewrite any section that fails one of these checks:

1. Can a director state the ask, cost, and recommendation after reading page one only?
2. Is there a genuine alternative, and is do nothing costed?
3. Is every size claim backed by a number, a range, or an explicit "we cannot quantify this because..."?
4. Are facts, assumptions, and judgements visibly separate?
5. Are the risks of the recommendation as clear as its benefits?
6. What are the eight hardest questions a prepared non-executive director would ask? Does the paper answer each one, or name it as an open point? List them in your working notes and fix the gaps in the paper, not in a verbal briefing.
7. Does implementation name an owner, a date, a success measure, and a report-back point?
8. Could the paper be read cold, three months later, and still make sense for the minutes and the audit trail?
9. Is anything material hidden in an appendix, a footnote, or a link?
10. Does the paper respect the reading load of the whole pack, not just this item?

Also check timing and process: papers should go out with the agenda early enough for directors to read and prepare. Late papers force a choice between a weak decision and a deferred one. Flag lateness to the sponsor instead of quietly tabling the paper on the day.

### 8. Rewrite rules for weak drafts

When given an existing draft, work in this order and explain each change briefly to the user:

1. Move the ask and resolution to page one
2. Cut background that does not change the decision
3. Replace size adjectives with numbers or ranges
4. Add the missing option, usually do nothing, or delete the straw option
5. Surface buried risks and state the tradeoff in the recommendation
6. Add owner, timeline, and report-back to the close
7. Move supporting data to appendices, keep material facts in the body

## Output Contract

Deliver:

- The full board paper in the structure in step 3, ready to paste into the board's house template
- A cover sheet block at the top
- A one-page executive lead that stands alone
- An options table or side-by-side option blocks using the same fields for each option
- A risk table with owner and residual level
- A draft resolution, one numbered line per separate decision
- A short author note after the paper, kept separate from it, listing: assumptions that need confirmation, numbers the user must supply, and any challenge-test gaps you could not close

Use the template in `assets/board-paper-template.md` as the starting skeleton. Use the paired drafts in `references/worked-examples.md` to judge tone and density.

If key facts are missing, do not fill them with plausible numbers. Leave a bracketed placeholder like [CONFIRM: total cost] and list it in the author note.

## Operating Rules

1. Accuracy first. Never invent figures, dates, votes, quotes, legal advice, or prior board decisions. Mark estimates as estimates and give the basis.
2. No selling. State disadvantages and residual risk for the recommended option as plainly as its benefits. A paper that reads like a pitch loses trust.
3. The board decides, management recommends. Keep that voice: "management recommends", "the Board is asked to", not "we will" before approval.
4. Plain language. Short sentences, active voice, no jargon without a one-line explanation. A director from outside the sector should follow the argument.
5. Confidentiality. Keep the classification on the cover sheet. Do not add names, personal data, or commercially sensitive figures the user did not provide.
6. One paper, one main ask. If a draft carries two unrelated decisions, split it or list the decisions separately and say why they are bundled.
7. Respect house style where the user provides it: template, length limit, currency, and resolution wording win over the defaults here.