openskills.info

Decision-Making Frameworks

Matching how much deliberation a decision gets to how much it would cost to get wrong.

A decision-making framework is a way of deciding how to decide. The most useful ones do very little clever work and a lot of triage: is this reversible or not, who actually owns it, what would change my mind, and how much deliberation is this worth. That last question is the one most teams skip, and it's where most of the waste is — reversible decisions get committee treatment while irreversible ones get made in a corridor. Frameworks don't produce right answers. They make your reasoning visible enough that someone can disagree with a specific part of it, and they let you tell later whether a bad outcome came from a bad decision or from luck.

Mostly Harmless

The Matrix Builder

Constructs a weighted scoring model of real sophistication — criteria, weights, a sensitivity analysis — and then adjusts the weights until the model recommends the option they walked in preferring. Each adjustment is individually justifiable and made with a straight face. The output is a decision that now carries the authority of arithmetic, and a spreadsheet nobody will interrogate because interrogating spreadsheets is tedious and it looks thorough. The preference was formed in about four seconds, some weeks earlier, and has now been laundered into a methodology.

The Reversibility Blind

Applies one deliberation setting to everything. Convenes three meetings and a written recommendation to choose a project management tool that could be swapped next quarter at a cost of an afternoon, then approves a two-year data migration in a corridor because the person asking seemed confident and it was nearly lunchtime. Both decisions received careful thought; the allocation of that thought was made by scheduling accident. Would describe themselves, accurately, as someone who takes decisions seriously.

The Option Hoarder

Keeps everything open. Declines to close down any avenue, on the reasonable-sounding grounds that more information will arrive later, and maintains four live possibilities well past the point where pursuing all four costs more than picking wrong. The decision is eventually made by a deadline, a departure, or an external event, and the Option Hoarder regards this as bad luck rather than as the predictable output of a strategy.

The Outcome Grader

Judges every past decision purely by how it turned out. The well-reasoned bet that lost is filed as a mistake and its author quietly marked down; the reckless call that happened to land becomes a case study and gets imitated. Over a few years this teaches everyone the actual lesson on offer — that visible downside is the thing to avoid — and the organization stops making good decisions with uncertain outcomes altogether.

Scenarios

You're asked to approve something that could be undone next month at almost no cost. What would you do — give it the same scrutiny as anything else with your name on it, or decide quickly and spend the saved time on the migration nobody has reviewed?

A decision you made last year turned out badly. What would you do — treat it as a mistake to learn from, or first ask whether you'd make the same call again given only what you knew at the time?

Your team has been "gathering more information" on a choice for six weeks. What would you do — keep gathering, or write down what specific piece of information would actually change the answer, and set a date?

Takeaway

  • Sort by reversibility first. Cheap-to-undo decisions deserve speed, not process.
  • Write down what would change your mind before you decide. It's the only way to tell later whether you were wrong or just unlucky.
  • Name the owner. A decision everyone is consulted on and nobody owns gets made by a deadline instead.
  • Judge past decisions on the reasoning available at the time, out loud, especially when the outcome was bad.
  • Watch for the framework being tuned to the answer. If the weights moved after you saw the result, you made the decision first.

Go deeper

Awesome Links