The minute that matters

A minute is evidence, and it is usually thin

I keep the firm’s decision archive. It holds 112 files between 2019 and 2026. Part of my work is reading old minutes with a question they were never written to answer: on what basis did they decide?

A standard minute answers who was present, what was proposed, who abstained and what carried. All of that is necessary. None of it is reasoning. Two years later the minute proves that a decision was taken and says nothing about whether it was taken well.

The two paragraphs that are missing

A minute becomes useful when it adds two things.

The first is the alternatives that were live at the vote, named, with one line each on why they lost. The second is the evidence the board did not have, and what it assumed instead.

That is two paragraphs, written while the room is still in it. It is the difference between a record and a receipt.

Write it in the room

Reconstruction fails, and it fails in a predictable direction. Someone drafts the minute weeks later from memory and a slide deck, and memory has already tidied the decision. The option rejected on cost becomes the option nobody took seriously, and the assumption the board made under pressure becomes a finding.

Nobody is lying. Recall works this way on everyone, including the people who write minutes for a living.

So we draft the reasoning paragraphs during the session and read them back before the item closes. An objection at that moment costs a minute of board time. The same objection at the appeal costs a legal review.

What 400 slides could not do

A water utility asked us to look at an investment programme running eleven years. The previous contractor had left 400 slides. Every number in them was traceable to a source.

Not one page named another shape the programme could have taken, or why this one won. The directors who approved it had moved on. The programme was defensible on arithmetic and indefensible on reasoning, and the reasoning was what the regulator asked about.

The test we run on our own files

Take a decision out of the archive. Give the file to someone who was not there. Ask two questions: what did the board believe would happen, and what would have changed its mind?

If the file answers both, the record works. If the reader has to telephone someone, you do not have a record. You have a pile of documents. Files we inherit usually fail this test, and they fail in the same place every time. The numbers are there. The alternatives are gone.

What the absence costs

Writing the reasoning costs a few minutes of a board meeting and no money at all.

Not writing it costs a rebuild: interviews with people who have left, emails read without their context, a narrative assembled long after the fact. The person assembling it usually has an interest in how it turns out. We have done that rebuild for clients. It is slow and expensive, and the result is always weaker than one honest paragraph written by someone who was in the room.