Start with the decision the board must make
A board paper exists to support a decision, a discussion, or an assurance item, and the paper should say which one in its opening lines. If the author cannot name the decision, the paper is not ready for review. The support desk should ask for that sentence before checking formatting or page count. A clear ask determines what evidence the rest of the paper needs.
Many packets contain several papers with different purposes, so the checklist should be applied to each one. A decision paper needs options and a recommendation, while an assurance paper needs evidence and exceptions. Mixing the two produces a document that informs without enabling action. The desk can catch that mismatch early by reading the first page of every paper.
Check the narrative before the layout
A polished layout can hide a weak argument, so review the story before adjusting fonts or spacing. The paper should explain the situation, the complication, the options, and the recommendation in a sequence a director can follow without prior context. If the reader has to reconstruct the logic from an appendix, the narrative needs work. The desk can fix a sentence order much faster than it can rebuild an argument.
Look for paragraphs that describe activity without stating a consequence. A sentence about a completed project should say what changed for the organization, not only what the team did. Cut repetition between the executive summary and the body, because duplication makes the paper longer without making it clearer. The desk's job is to protect the reader's attention, not to fill a template.
Test the numbers against one source of truth
Board papers often contain numbers drawn from different systems and different dates. The checklist should require a named source for each table and a note on when the data was extracted. If two tables disagree, the desk should resolve the difference before submission rather than letting a director find it in the meeting. One reconciled figure is worth more than a dozen polished tables.
Ask the author to reconcile the summary figures with the detailed appendix. Rounding differences are acceptable, but unexplained gaps are not. The desk should also check that the units, currencies, and time periods are consistent across the document. A single mismatched unit can undermine confidence in an otherwise strong paper.
Review risk and mitigation in plain language
A risk section should describe what could happen, how likely it is, and what the organization would do about it. It should not be a list of generic categories copied from a template. The desk should check that each material risk has an owner and a mitigation that someone can actually execute. A risk without an owner is a warning, not a control.
Look for the risk that the paper avoids discussing. If the recommendation depends on a vendor, a hire, or a regulatory approval, the paper should state the dependency and the fallback. Directors are better served by an honest constraint than by a confident plan with no downside. The desk can ask the author one direct question about the biggest thing that could go wrong.
Confirm the ask, the owner, and the date
Every decision paper should state exactly what the board is being asked to approve, who will execute it, and when the work begins. The ask should be specific enough to be minuted, such as approving a budget increase rather than noting a direction. The owner should be a named role, and the date should be a calendar date rather than a quarter. A minutable ask prevents the meeting from ending with a direction that nobody can execute.
Check that the paper does not ask for approval of something that has already happened. Retrospective approval is sometimes necessary, but it should be labeled clearly and accompanied by an explanation. The desk should also confirm that the requested authority matches the board's mandate. A paper that asks for the wrong approval wastes the meeting and delays the work.
Run a red-team read for the chair
Before the packet is frozen, one person who did not write the paper should read it as a skeptical director. The reader should look for unclear recommendations, unsupported claims, and questions that the paper does not answer. This review is faster than a full edit because it focuses on the three or four things most likely to derail the discussion. The list of unanswered questions becomes the agenda for the author's final revision.
The red-team reader should not rewrite the paper. The reader should list the questions a director would ask and send them to the author. The author decides how to address them, and the desk confirms that each material question is answered somewhere in the packet. This step catches the gaps that a formatting review cannot see.
Freeze the packet and log late changes
Set a freeze time that gives the chair and the board enough time to read the packet. After the freeze, every change should be logged with the reason, the person who approved it, and the version it replaced. A late change without a log creates a risk that directors are reading different documents in the same meeting. The version label is the simplest way to keep the room on the same page.
The support desk should distribute the final packet with a clear version label and a short note about any changes since the previous distribution. Keep the approved version and the change log together. For related guidance, see our article on building a decision log that survives turnover and our executive support services. The GAO Green Book provides a useful standard for documentation and internal control.
Key takeaways
- Require each paper to state whether it is for decision, discussion, or assurance.
- Review the narrative before the layout, and cut activity descriptions that do not state a consequence.
- Reconcile every summary figure with one named source and one extraction date.
- Confirm the specific ask, the accountable owner, and the start date before the freeze.
- Use a skeptical reader, a freeze time, and a change log to protect the final packet.