Decision-making & prioritisation
OODA Loop
A four-phase decision cycle, Observe, Orient, Decide, Act, developed by fighter pilot John Boyd. Advantage goes to the side that reads a changing situation faster and more accurately, with orientation, not raw speed, doing most of the work.
Also known as Boyd cycle, Observe-Orient-Decide-Act loop. First set out by John R. Boyd in 1976; the primary source is cited in full below.
Where this is contested
Boyd never drew the simple four-step circle; it is a later popularisation. His only sketch of the loop, in 'The Essence of Winning and Losing' (1995-96), is far richer, with Orient as the governing element and an implicit guidance and control path running straight from Orient to Act.
- Format
- Process / loop
- Level
- Business unit · Team
- Best for
- Position against competitors · Evaluate options · Plan execution
- Decision stage
- Diagnose · Decide · Execute
- Difficulty
- Intermediate
- Time to apply
- An hour to map a team's current cycle honestly; embedding the loop as an operating rhythm takes weeks.
Plate · The model
The components
Observe
Take in unfolding circumstances, outside information and the results of your own actions. Observation feeds orientation and is shaped by it in turn; you tend to see what your current model prepares you to see.
Signals of strength
Fresh, unfiltered data reaches decision-makers quickly · Mismatches between forecast and outcome are noticed rather than explained away · Competitor and customer behaviour is watched directly, not via summaries alone
Orient
Make sense of what is observed by filtering it through culture, experience, analysis and synthesis. Boyd's sketch shows Orient as the richest box in the loop and the one that shapes all the others. It is where old mental models are destroyed and new ones built.
Signals of strength
The team can state the assumptions behind its current reading of the situation · Alternative interpretations are generated, not just one official view · Orientation is updated when actions produce surprising results
Decide
Select a course of action as a hypothesis about what will work given the current orientation. In a well-drilled organisation many actions bypass explicit decision entirely, flowing straight from shared orientation.
Signals of strength
Decisions are made at the lowest competent level · Choices are framed as testable rather than final · Decision rights are clear enough that the loop does not queue behind one person
Act
Execute the choice and interact with the environment. Action tests the hypothesis; its consequences become the next cycle's observations. Speed of action matters, but only if orientation has made the action roughly right.
Signals of strength
Actions produce feedback that is actually collected · The gap between deciding and doing is short · Repeated cycles show learning rather than the same move replayed
When it earns its keep
- You are competing head-to-head in a fast-moving environment where the side that senses and responds to change first wins, for example pricing, incident response or a contested bid.
- Your organisation reacts to events slowly and you want to diagnose where the cycle stalls: in gathering information, in making sense of it, in deciding, or in acting.
- You suspect your team's mental models are stale and its decisions are technically fast but consistently wrong, which points at orientation rather than process.
- You are designing an operating rhythm and want decisions treated as hypotheses whose results feed straight back into the next cycle.
And when it doesn't
- The decision is a one-off, high-stakes commitment with time to analyse. A decision tree or cost-benefit analysis will serve better than a tempo framework.
- The environment is stable and the work is repeatable. A control loop like PDCA, which is built for incremental improvement, fits better than a competitive cycle.
- You are reaching for it as a slogan. 'Get inside their OODA loop' used as a call to simply move faster misreads Boyd; without work on orientation the loop adds nothing.
- The problem is disagreement over goals rather than speed of response. OODA assumes you know what winning looks like.
How to run it
Before starting, gather the inputs the analysis depends on:
- Live feeds of what is actually happening: market signals, competitor moves, customer behaviour, operational data, not last quarter's report.
- An honest inventory of the team's orientation: the assumptions, experience, culture and models through which it interprets what it sees.
- Clarity on who can decide and act at each level, since a loop that routes every decision upwards cannot be fast.
- A way of capturing what each action taught you, so the next cycle starts better informed.
- 1
Set up observation
Decide what the team watches and how often. Observation is not passive data collection; it includes noticing the mismatch between what you expected and what unfolded, which is the raw material for the rest of the loop.
- 2
Invest in orientation
Orient is where observations are filtered through genetic heritage, culture, previous experience and analysis to form a picture of the situation. Boyd treated this as the schwerpunkt, the focal point of the whole scheme. Most organisational failures blamed on slow decisions are really failures of stale orientation.
- 3
Decide as hypothesis
Choose a course of action explicitly framed as a test of your current picture. In Boyd's sketch the decide step is literally labelled 'hypothesis'; the point of deciding is partly to find out whether your orientation is right.
- 4
Act and feed the result back
Carry out the decision and observe what the action changes. Action is the test of the hypothesis, and its results flow back into observation and orientation. Well-oriented teams often act directly from orientation without a deliberate decide step, which Boyd called implicit guidance and control.
- 5
Work on tempo
Once the loop runs honestly, compress it. Operating at a faster and more accurate rhythm than a competitor makes their picture of the situation progressively wrong, which Boyd argued generates confusion and paralysis on the other side.
Reading the result
A running operating rhythm rather than a document: a clear picture of where your decision cycle is slow or misoriented, decisions framed as hypotheses, and a feedback discipline that improves the organisation's read of its environment with every pass.
- Find the binding constraint in your loop. Teams usually assume they observe too little or decide too slowly; more often they orient badly, filtering good data through stale assumptions.
- Judge advantage relatively. The question is never whether your cycle is fast, it is whether it is faster and better oriented than the competitor's or the situation's rate of change.
- Watch for implicit guidance and control emerging. When shared orientation is strong enough that people act correctly without waiting for decisions, the loop is working as Boyd intended.
A worked example
An online travel agency's pricing team rebuilds its repricing cycle
The hotel pricing team at a UK online travel agency is losing bookings to two rival agencies that seem to undercut its rates within hours of any change. The team repriced weekly from a batch report. Before buying a faster repricing tool, the head of pricing runs the operation through the OODA loop to find out where the cycle actually fails.
- Observe
- Weak but fixable. Competitor rates were scraped daily but only surfaced in a Monday report, so the team was acting on observations up to a week old. Conversion data by rate position existed in the analytics platform but nobody looked at it. The fix was cheap: a live dashboard, not a new tool.
- Orient
- The real problem. The team's shared model said price rank drives conversion, so every cycle ended in matching the cheapest rival. The conversion data contradicted this for packages and late bookings, where cancellation terms and payment options mattered more than a three-pound gap. Orientation had not been updated since the team was formed.
- Decide
- Slow for structural reasons. Any rate move beyond two per cent needed sign-off from the commercial director, who reviewed requests twice a week. Reframed as hypothesis-testing within guardrails, repricing decisions inside an agreed band were delegated to the analysts running the loop.
- Act
- Adequate once decisions arrived: rate changes propagated to the site within the hour. But results of each change were never systematically compared with the expectation behind it, so acting taught the team nothing. A one-line prediction logged with each change closed the feedback gap.
The read. The team assumed it had a speed problem and nearly bought tooling to observe and act faster. The loop showed the binding constraint was orientation: a wrong shared model of what drives conversion, plus decision rights that queued the cycle behind one director. After delegating in-band repricing and testing the price-rank assumption segment by segment, cycle time fell from a week to a day and margin improved on late bookings without matching rivals' cuts. Faster software would have executed the wrong model more quickly.
Pitfalls
- Treating the loop as a simple sequential circle. Boyd's own sketch is dense with feedback and forward paths, and orientation shapes every other phase; the four-step ring is a teaching aid, not the theory.
- Reducing it to 'be faster'. Cycling quickly with a wrong orientation just produces confident errors at pace.
- Skipping the orient work because it is uncomfortable. Examining culture, assumptions and past experience is the hard centre of the framework, and the part most business applications quietly drop.
- Centralising decisions while preaching tempo. A loop that routes through a weekly approval meeting has a fixed floor on its speed regardless of how good the observation is.
- Using it where there is no adversary and no time pressure, where its competitive logic adds ceremony without insight.
What the critics say
Osinga's full-length study argues the popular rapid-OODA idea is a caricature. Boyd's actual theory concerns adaptation, learning and the destruction and creation of mental models under uncertainty, and reading the loop as a speed formula strips out nearly everything he meant by it.
Osinga, F. P. B. (2007) Science, Strategy and War: The Strategic Theory of John Boyd. Abingdon: Routledge.
Richards, who drew Boyd's original graphics, argues the widely reproduced four-box circle does not appear anywhere in Boyd's work and that the real loop, with implicit guidance and control running from Orient to Act, describes how trained organisations mostly bypass explicit decision-making. Much business writing on OODA therefore describes a diagram Boyd never endorsed.
Richards, C. (2020) 'Boyd's OODA Loop', Necesse, 5(1), pp. 142-165.
The framework offers little concrete guidance on how to improve orientation, its own centre of gravity. It diagnoses that mental models matter but supplies no method for eliciting or testing them, so practitioners must import tools from elsewhere to act on the insight.
Sources and further reading
- Boyd, J. R. (1976) 'Destruction and Creation', unpublished essay, September 1976.
- Boyd, J. R. (1995-96) 'The Essence of Winning and Losing', briefing slides, final revision January 1996. ↗
- Coram, R. (2002) Boyd: The Fighter Pilot Who Changed the Art of War. New York: Little, Brown.
- Osinga, F. P. B. (2007) Science, Strategy and War: The Strategic Theory of John Boyd. Abingdon: Routledge.