John Boyd was a fighter pilot and military theorist who wanted to explain why a plane that was slower and less maneuverable on paper kept winning. His answer was that the pilot who could complete the cycle of observing, orienting, deciding and acting faster than his opponent was effectively fighting someone who was responding to a situation that no longer existed.

The model transferred to business, mostly in a degraded form that reads it as "move fast". Boyd's actual argument is more specific and more useful.

The four steps, and which one matters

Observe. Take in what is happening. Limited by what you are set up to notice — and most organizations observe what they already measure, which is a narrower thing.

Orient. Interpret it. Fit the observation into a picture of how things work.

Decide. Choose a response.

Act. Do it, which changes the situation and produces new observations.

Boyd's view was that orientation is the whole game. It is where previous experience, training, culture and expectation do their work, and it is the step that determines what the observation means. Two people can see the same number and orient to it completely differently — one sees a bad month, the other sees the first sign of a trend.

Which is why the reading of OODA as speed is a mistranslation. A fast loop on a bad orientation produces confident wrong answers more frequently. The competitive advantage Boyd described came from reorienting — being able to drop a model of the situation that had stopped fitting — and that is a different and harder skill than being quick.

Why reorienting is hard

Because the orientation is what makes the observations legible in the first place. Dropping it means losing the ability to interpret anything for a while, which is unpleasant enough that people mostly do not, and instead absorb contrary evidence into the model they have.

This is the same structure Kuhn described in science, where anomalies accumulate for a long time before the frame gives way — see paradigm shift. The business version is a firm that keeps explaining each lost deal individually rather than concluding that the market has changed.

What it is good for

Spotting where the loop is slow. Most organizations are slow at decide, not at act. Work sits waiting on an approval, on a meeting, on a person who is away. That is a batch size problem wearing a strategy costume.

Noticing you are observing the wrong thing. If the numbers on the dashboard have stopped changing when the situation changes, the observation step is broken, and no amount of speed in the other three will help.

Deciding how fast to decide. A loop that can be run again cheaply should be run quickly and imprecisely; one that cannot should be slower. That question belongs to reversible decisions; OODA is the reason to ask it.

The honest limit

Boyd was describing aerial combat and then conflict generally — domains with an opponent who is actively trying to defeat you, and where the situation changes primarily because of what the two of you do.

Most business situations are not that. Customers are not opponents, markets are not adversaries, and a firm obsessed with out-cycling a competitor is frequently running a fast loop against something that is not reacting to it at all. Where there is a genuine adversary — a competitive tender, a negotiation, a market with few players watching each other — the model holds well. Elsewhere, the useful residue is narrower and still worth having: notice faster, be willing to change your mind about what it means, and match the pace of the decision to whether you can take it back.