What a board actually reads

The order of reading is not the order of writing

A director opens the pack the evening before. She reads the recommendation first. Then she looks for the number that would embarrass her if it turned out to be wrong. Then she looks for the paragraph where the advisers hedge, because that is usually where the weakness is. The introduction, written first and revised most often, gets read last or not at all.

She is triaging: one evening, a full agenda, legal responsibility for the outcome.

What survives the skim

Four things survive. The decision stated as a sentence, and the number it rests on. The alternative that was rejected, with the reason it lost. The condition that would make the recommendation wrong.

Everything else is support. Support matters, because a regulator may read it when the file is reopened. The room decides on the four.

When those four things are scattered through the pack, the board reconstructs them in the meeting, out loud and badly. That work eats the time that should have gone on the choice.

The board that asked for eight pages

An insurer’s board once asked us for eight pages. The first reaction inside the team was that we would have to thin the analysis to fit.

The opposite happened. Eight pages forced us to decide which of our scenarios actually changed the vote. Some of them did not. We had been carrying them because they were interesting, and interesting is not a reason. The short pack took longer to write and was harder to argue with, because we could defend every line of it.

The forty minute test

A board gives an item about forty minutes, if the chair is firm. That is the session the material is written for, and everything in the pack competes for it. Set the pack against that length before it goes out. If a third of the session would go on context the board already has, the pack has failed. The quality of the annex will not save it.

We write the handover document against that session, and we write the annex separately, for a different reader on a different day. Two documents, one record. Mixing them produces material that is too long for the meeting and too thin for the file.

Write the first page last

The first page is the only page with guaranteed readers. It carries the four things above, plus the date the answer should be looked at again.

That page cannot be written at the start of the work. At the start you do not know which uncertainty is load bearing. Write it at the end. Then read the rest of the pack and ask what each section does that the first page has not already done. Sections that fail that question go to the annex. Some of them go nowhere.