Running a discussion that can close

One: sort the proposal document

A proposal that sorts its own claims is very hard to answer badly.

What we think the evidence shows, and what we would look at: new starters reach independent delivery later when they begin remotely. We have the cohort data and have attached it.

What we think matters, which is a judgement and not a finding: that an organisation is better when its people are regularly in the same place.

What it costs and who bears it: the building, which we hold to 2031; and the travel, which falls on individuals and ranges from eighteen minutes to an hour and twenty-five each way.

Any reply to that has to engage with three separate things, and it is immediately obvious when a reply has only engaged with one.

Two: the forty-second opening

Said once, near the start, by whoever is chairing:

Some of what we say today will be about what the evidence shows, some about what we think matters, and some about who pays for what. Those are three different conversations and they do not answer each other. It would help if we could say which one we are having.

Not a framework and not a process. A request, made once, that gives everybody in the room permission to do the thing.

Three: take the notes in three columns

The highest-leverage item, and it is usually done by the most junior person present.

Nobody has to be told this is happening. The notes come back in three columns, and the next meeting starts from a document in which the sorting has already been done by the record rather than by a person.

Four: close each column differently

At the end, three different closing moves:

• Column one gets assigned. Somebody goes and looks, by a date.
• Column two gets stated as a disagreement, in both parties' words, and either traded or explicitly overruled.
• Column three gets designed around: notice periods, which days, how long the arrangement holds.

That is a meeting that can produce something. The alternative — a single list of what everybody said, followed by a decision — is what produced eleven weeks and an announcement on this record.

What this does not do

It does not make anybody agree, and it will not stop a decision that was going to be made anyway.

What it changes is whether the decision is made in the presence of the costs, whether the empirical questions get answered or just argued, and whether the people overruled were overruled on the thing they actually said. On this record none of the three was true, and all three were available for the price of three columns in a notebook.

What to do if you are not the chair

Most people reading this will not be running the meeting, and three of the four items are still available.

The three columns can be kept privately, by anybody, in a notebook. They are useful even if nobody else ever sees them, because they tell you which claims are worth pursuing and which are going nowhere.

The forty-second opening can be offered rather than declared: "It might help to separate the bits that are evidence from the bits that are judgement — would that be useful?" Asked as a question, from any position in the room, it is usually accepted.

The three different closes can be suggested one at a time: "Could somebody go and look at that one before next time?" is a sentence anybody can say, and it is the single most useful intervention available to a junior person in these discussions.

The thing that only a chair can do

One item does require the chair, and it is the one that matters most: closing the value column honestly.

Only the person deciding can say "we disagree about this, I have understood it, and I am deciding against it". Nobody else can do that on their behalf, and no amount of good process substitutes for it.

Which means that if you are the chair, that sentence is the specific thing you owe the discussion, and it is the one most often omitted.