Contents
The OODA Loop is a decision-cycling framework for environments where speed of adaptation matters more than perfection of analysis. Use it when the situation is shifting faster than your planning horizon — and the cost of a late decision exceeds the cost of a wrong one.
Section 1
What This Tool Does
John Boyd never published a book. He never held a tenured position. He retired from the United States Air Force as a colonel — not a general — and spent the last decades of his life giving marathon briefings to anyone in the Pentagon who would listen, scribbling on overhead projector slides for six, eight, sometimes twelve hours straight. He died in 1997 with almost no formal written legacy. Yet his core idea — a four-phase cycle of Observe, Orient, Decide, Act — has shaped military doctrine, business strategy, startup methodology, and cybersecurity thinking for half a century. The gap between Boyd's obscurity and his influence is itself a lesson in how ideas propagate: not through publication, but through operational superiority.
Boyd developed the OODA Loop to explain why American F-86 Sabre pilots achieved a 10:1 kill ratio against technically superior Soviet MiG-15s during the Korean War. The MiG was faster, climbed better, and could operate at higher altitudes. On paper, it should have dominated. But the F-86 had a hydraulic flight control system and a bubble canopy that gave pilots faster stick response and wider visibility. Boyd's insight was that the advantage wasn't aerodynamic — it was temporal. F-86 pilots could observe the battlespace more quickly, reorient to changing conditions faster, make decisions with less lag, and act before MiG pilots had finished processing what was happening. The F-86 didn't win by being better. It won by being faster through the decision cycle.
The loop has four phases, but they are not created equal. Orient is the critical phase — the one Boyd called "the big O" — because it's where your mental model of reality gets updated, and a flawed mental model will corrupt every decision downstream regardless of how fast you cycle. Observe gathers data. Orient interprets it through the filters of cultural traditions, previous experience, genetic heritage, and new information — Boyd's list, not a metaphor. Decide selects a course of action. Act executes it. Then the loop restarts, because your action has changed the environment, which means your previous observations are already stale.
Most people misunderstand the OODA Loop as "just go fast." That's the cartoon version. Boyd's actual framework is about operating inside your adversary's decision cycle — completing your loop before they complete theirs, so that by the time they act on their observations, the situation has already changed because of your action. The result isn't just speed. It's disorientation. Your opponent starts responding to a reality that no longer exists. They make decisions based on obsolete observations. Their actions become increasingly disconnected from the actual environment. This is what Boyd called "getting inside someone's OODA Loop," and it applies to market competition as cleanly as it applies to aerial combat.
The framework's deepest contribution isn't the loop itself but the emphasis on orientation as a bottleneck. Most organisations invest heavily in observation (data collection, market research, competitive intelligence) and agonise over decisions (committees, approvals, consensus). Almost none invest in improving the quality and speed of orientation — the interpretive layer where raw data becomes situational understanding. A company with perfect data and a broken mental model will consistently lose to a company with imperfect data and a superior ability to make sense of what it sees.
Section 2
How to Use It — Step by Step
Instructions on the left. Worked example — a Series B fintech company discovering that a major competitor has just launched a free version of its core product — on the right.
Step 1 — Observe
Gather raw signals from the environment without filtering
Cast wide. You're collecting data, not interpreting it. Pull from multiple channels: customer behaviour metrics, competitor announcements, sales team anecdotes, social media sentiment, support ticket patterns, industry news. The discipline here is to resist premature interpretation. "Our competitor launched a free tier" is an observation. "Our competitor is trying to destroy us" is an orientation — and jumping to it too early will blind you to alternative explanations. Collect signals that might contradict your initial read, not just ones that confirm it. Time-box this phase. In a fast-moving situation, 24–48 hours maximum. In a crisis, 2–4 hours.
Worked example
Fintech competitor launches free product
The team collects: (1) Competitor's press release announcing a free version of their payment analytics tool. (2) Twitter/X chatter — mostly positive, some scepticism about data limits. (3) Internal data: no spike in churn yet, but three enterprise prospects have paused contract negotiations "to evaluate alternatives." (4) Sales team reports that two existing customers asked about price matching. (5) Competitor's free tier has a 10,000-transaction limit — well below enterprise needs but sufficient for SMBs, which represent 40% of the fintech's revenue. (6) A product review site notes the free tier lacks API access and custom reporting.
Step 2 — Orient
Interpret the signals through your mental models — then challenge those models
This is the phase that separates competent operators from exceptional ones. Take the raw observations and ask: What do these signals mean? What pattern are they forming? What assumptions am I bringing to this interpretation, and which of those assumptions might be wrong? Boyd identified five filters that shape orientation: cultural traditions, previous experience, genetic heritage (temperament), new information, and the ability to analyse and synthesise. In practice, the dangerous filters are previous experience and cultural tradition — they create pattern-matching shortcuts that work until the environment changes. Force yourself to generate at least two competing interpretations of the same data. If you can only see one explanation, your orientation is locked.
Worked example
Two competing orientations
Interpretation A — Existential threat: The competitor is subsidising a free tier to capture SMB market share, planning to upsell to paid plans later. This is a classic land-and-expand play that will erode the fintech's SMB base within 6–12 months. Urgent response required.
Interpretation B — Defensive move: The competitor's core enterprise business is slowing (their last earnings showed flat growth), and the free tier is a lead-generation tactic, not a sustainable product. The 10,000-transaction cap and missing API access suggest they built this quickly, possibly cannibalising their own paid SMB tier. This may hurt them more than it hurts us.
The team notes that their instinct — shaped by previous experience losing a client to a free competitor in 2021 — is pulling hard toward Interpretation A. They flag this bias explicitly. They assign one person to build the strongest possible case for Interpretation B before any decision is made.
Step 3 — Decide
Select a course of action — biased toward reversibility and speed
You will never have complete information. The OODA Loop's design assumes this. The decision doesn't need to be optimal; it needs to be directionally correct and fast enough to stay inside the competitor's cycle. Favour reversible actions over irreversible ones. Favour actions that generate new information (which feeds back into the next Observe phase) over actions that merely commit resources. If you're choosing between two options of roughly equal merit, pick the one that preserves more future optionality. And decide explicitly — "we're going to wait and see" is a decision, but only if it's conscious and time-bounded. Undeclared inaction is not a decision; it's drift.
Worked example
Choosing a response
The team decides on a three-part response, sequenced by reversibility:
Immediate (this week): Sales team gets a competitive battle card with specific feature gaps in the free tier (no API, transaction cap, no custom reporting). This is cheap, reversible, and addresses the three paused prospects directly.
Short-term (next 30 days): Launch a "free trial" of the full product for SMB prospects — not a permanent free tier, but 60 days of unrestricted access. This generates data on whether SMBs actually convert at the end of a trial, informing the next loop.
Conditional (hold unless triggered): If churn among SMB customers exceeds 5% in the next quarter, introduce a permanent free tier with the 25,000-transaction cap that the competitor can't match. This is the irreversible option — held in reserve, not deployed yet.
Step 4 — Act
Execute and immediately begin observing the results
Speed of execution matters as much as speed of decision. A decision that takes three days to implement has a three-day lag built into your loop. Assign clear owners, set tight timelines, and — critically — define what you'll observe in the next cycle. What signals will tell you whether your action is working? What would make you change course? The act phase isn't the end of the loop; it's the beginning of the next observation phase. Build the feedback mechanism before you execute, not after.
Worked example
Execution with embedded feedback
Battle cards ship to sales by Friday. The VP of Sales will track: (a) whether the three paused prospects re-engage within 10 business days, and (b) whether any new prospects cite the competitor's free tier as a reason for hesitation.
The free trial launches in two weeks. Product analytics will track: (a) SMB sign-up volume, (b) feature usage during the trial, (c) conversion rate at day 60, and (d) whether trial users who don't convert cite price or feature gaps.
The team schedules a 90-minute OODA review in 30 days. New observations will include all of the above plus: competitor's next moves, customer churn data, and any changes in the competitor's free tier (feature additions, cap changes). The loop restarts.
Section 3
When It Works Best
✓
Ideal Conditions for the OODA Loop
| Dimension | Best fit |
|---|---|
| Environment | Fast-moving, adversarial, or rapidly shifting contexts where the situation changes between planning cycles. Competitive markets, crisis response, product launches into contested categories, cybersecurity incidents. The tool's value scales with the rate of environmental change — the faster things move, the more you need it. |
| Information quality | Incomplete, ambiguous, or contradictory information. If you have perfect data and unlimited time to analyse it, you don't need the OODA Loop — you need a Decision Matrix. The loop is designed for situations where waiting for certainty means losing the window for action entirely. |
| Decision reversibility | Most powerful when the available actions are partially or fully reversible. The loop assumes you'll course-correct on the next cycle, which only works if your previous action hasn't locked you into an irreversible commitment. Pair with the Reversible vs. Irreversible Decisions framework to classify your options before choosing. |
| Competitive dynamics | Situations with an identifiable adversary or competitor whose actions directly affect your outcomes. The "getting inside their loop" concept requires someone whose loop you're trying to outpace. In non-competitive contexts (internal process improvement, personal decisions), the loop still works but loses its most distinctive feature. |
| Organisational agility | Teams that can actually execute quickly. The OODA Loop is useless in organisations where decisions require six layers of approval, because the Act phase takes so long that the environment has changed twice before anything happens. The loop exposes — and punishes — bureaucratic drag. |
| Leader temperament | Decision-makers comfortable with ambiguity and willing to act on 60–70% confidence. If your culture demands 95% certainty before acting, the OODA Loop will feel reckless. It isn't — but it requires a genuine tolerance for being wrong on individual cycles in exchange for being faster across many cycles. |
Section 4
When It Breaks Down
⚠
Failure Modes
| Failure pattern | What goes wrong | What to use instead |
|---|---|---|
| Speed fetishism | Teams interpret the OODA Loop as "decide faster" and skip the Orient phase entirely. They observe a signal, jump to a decision, and act — cycling rapidly through a loop that's missing its most important component. Fast decisions based on unexamined mental models produce confident, rapid failure. | Ladder of Inference to slow down the interpretive step; force explicit articulation of assumptions before deciding |
| Irreversible decisions treated as reversible | The loop's bias toward speed and iteration assumes you can course-correct. When the action is irreversible — a major acquisition, a public product pivot, a key hire — cycling fast through OODA can produce catastrophic commitments made on insufficient analysis. Not every decision deserves the same cycle speed. | Reversible vs. Irreversible Decisions to classify the stakes; Pre-Mortem for high-consequence choices |
| Locked orientation | The Orient phase is supposed to update your mental model. But if the team's cultural assumptions or previous experiences are too strong, orientation becomes confirmation bias with a fancy name. You "observe" only data that fits your existing worldview and "orient" by confirming what you already believed. The loop spins, but the model never changes. | Reframing to challenge the problem definition; assign a designated dissenter to argue the opposing orientation |
| Solo operator bias | Boyd designed the loop for fighter pilots — individual decision-makers with total authority. In organisations, the Observe and Orient phases often require input from multiple people with different vantage points, and the Decide phase may require buy-in from stakeholders. Running OODA as a solo exercise in a team context produces blind spots in observation and brittle decisions that lack organisational support. | Six Thinking Hats for structured multi-perspective orientation; Delphi Method for distributed observation |
| Stable environments | When the environment isn't changing rapidly, the OODA Loop's emphasis on speed creates unnecessary churn. Cycling through observe-orient-decide-act every week in a market that shifts quarterly produces organisational whiplash — constant pivots that confuse teams, exhaust resources, and erode strategic coherence. | Scenario Planning for slower-moving strategic decisions; Cost-Benefit Analysis when you have time to be thorough |
| Observation without synthesis | Teams become excellent at collecting signals — dashboards, alerts, competitive intelligence feeds — but never synthesise them into a coherent picture. The Observe phase balloons with data while the Orient phase starves. More information doesn't help if nobody is making sense of it. This is the most common corporate failure mode: drowning in data, starving for insight. | Issue Trees to structure the synthesis; First Principles Thinking to cut through noise to fundamentals |
The most dangerous failure mode is locked orientation, because it's invisible from the inside. Speed fetishism is obvious — you can see a team skipping analysis. Irreversible decisions are identifiable — you can flag them before they're made. But a team with a locked mental model feels like it's doing everything right. They're observing diligently. They're discussing what the data means. They're deciding and acting quickly. The loop is spinning. The problem is that the Orient phase has become a rubber stamp for pre-existing beliefs. Every new observation gets absorbed into the same interpretation. Disconfirming evidence gets explained away or simply not collected.
The protection is structural, not motivational. You can't just tell people to "be open-minded." You need a mechanism. Boyd's own approach was adversarial: he would argue the opposing position with ferocious intensity, forcing his audience to defend their orientation against a skilled attacker. In a business context, the equivalent is assigning someone — rotating the role each cycle — to build the strongest possible case for the interpretation the team is not pursuing. If that person can't be refuted with data, your orientation might be wrong.
Section 5
Visual Explanation
OODA Loop — populated with the fintech competitive response example. Orient (the 'big O') is shown larger to reflect its disproportionate importance in Boyd's framework.
Section 6
Pairs With
The OODA Loop is a meta-framework — it tells you when and how fast to think, but not how to think within each phase. Pairing it with the right tools at each stage is what turns rapid cycling into rapid learning.
Use before
Cynefin Framework
Before entering the loop, classify the domain. Is this situation chaotic (act first, then sense), complex (probe-sense-respond), complicated (sense-analyse-respond), or obvious? The OODA Loop works best in complex and complicated domains. In chaotic situations, you may need to collapse the loop to Act-Observe-Orient — skipping the Decide phase entirely and just stabilising.
Use before
Reversible vs. Irreversible Decisions
The OODA Loop's bias toward speed assumes course-correction is possible. Before deciding, classify the action: is it a one-way door or a two-way door? One-way doors deserve a slower, more deliberate loop. Two-way doors should be cycled through as fast as organisationally possible.
Use after
Pre-Mortem
After the Decide phase but before Act, run a rapid pre-mortem on high-stakes decisions: "It's 90 days from now and this action failed spectacularly. What happened?" This takes 15 minutes and catches failure modes that the Orient phase may have missed — without slowing the loop significantly.
Use after
[Second-Order Thinking](/mental-models/second-order-thinking)
After acting, before the next Observe phase, ask: "What will our competitor do in response to our action?" Second-order thinking anticipates the next iteration of the adversary's loop, letting you pre-position your observation for signals you expect to see rather than waiting passively.
Mental model
Ladder of Inference
The Ladder of Inference maps exactly what happens inside the Orient phase: you select data from observations, add meaning, make assumptions, draw conclusions, and adopt beliefs. Making this ladder visible — especially the rungs you're skipping — is the best defence against locked orientation.
Mental model
Confidence Determines Speed vs [Quality](/mental-models/quality)
A direct complement. When confidence is high and stakes are low, cycle the OODA Loop fast and bias toward action. When confidence is low and stakes are high, slow the loop down and invest more in the Orient phase. The two frameworks together create a dynamic throttle for decision speed.
Section 7
Real-World Application
Apple vs. Samsung — the smartphone patent wars and product cadence, 2011–2014
The scenario
Between 2011 and 2014, Apple and Samsung fought the most expensive patent litigation in consumer electronics history — over $2 billion in claims across courts in the US, South Korea, Japan, Germany, and the UK. But the legal battle was a sideshow. The real competition was happening in product cycles. Samsung was releasing new smartphone models at roughly three times Apple's cadence: multiple Galaxy variants per year across price points, screen sizes, and feature sets, while Apple released one iPhone annually. Analysts framed this as a weakness for Apple. It wasn't. Both companies were running OODA Loops — but at different speeds and with fundamentally different orientations.
How the tool appliedSamsung's loop was optimised for observation breadth and action speed. They watched what features gained traction across the entire Android ecosystem — large screens, stylus input, removable batteries, expandable storage — and rapidly shipped products that incorporated whatever was working. Their Orient phase was market-reactive: what do consumers seem to want right now? Their Decide-Act cycle was extraordinarily fast, enabled by vertical integration in display and semiconductor manufacturing. They could go from "large screens are trending" to "Galaxy Note on shelves" in months.
Apple's loop was slower but oriented differently. Their Observe phase focused less on competitor features and more on component technology maturation, supply chain readiness, and user behaviour data from the existing installed base. Their Orient phase filtered observations through a design philosophy that prioritised integration over feature checklists. When Apple finally moved to larger screens with the iPhone 6 in 2014, they didn't just make a bigger phone — they redesigned the entire interaction model, renegotiated display supply contracts to secure volume at a specific quality threshold, and timed the launch to coincide with iOS 8's adaptive layout system.
What it surfacedThe iPhone 6 and 6 Plus became the best-selling smartphones in history at that point, with Apple selling over 74 million units in Q1 FY2015 alone. Samsung's smartphone profits, meanwhile, declined roughly 60% year-over-year in Q3 2014. Samsung had cycled faster, but Apple's orientation was superior — they understood that the large-screen transition wasn't just a hardware change but a platform transition that required software, services, and supply chain alignment. Samsung won the feature race. Apple won the profit race. The OODA Loop explains why: Samsung's rapid cycling produced many adequate responses. Apple's slower, more deeply oriented cycling produced one devastating response.
The non-obvious factorBoyd would have recognised this immediately. The OODA Loop doesn't say "fastest wins." It says the entity that completes the most effective cycle wins — and effectiveness is determined by the Orient phase, not the Act phase. Apple's willingness to appear slow, to let Samsung "win" the large-screen narrative for two years, was itself an orientation decision: they judged that the cost of being late was lower than the cost of being wrong. Samsung's rapid cycling, by contrast, created an illusion of strategic initiative that masked a reactive orientation. They were fast followers moving quickly — which is different from leading. The lesson for founders: cycle speed is a variable you control, not a constant you maximise. The right speed depends on what your Orient phase tells you about the stakes.
Section 8
Analyst's Take
Faster Than Normal — Editorial View
The OODA Loop endures because it solves a problem that no amount of data infrastructure or analytical sophistication can solve on its own: the gap between knowing and doing. Every organisation I've worked with has more information than it can process and more analysis than it can act on. The bottleneck is never observation. It's the willingness to form an interpretation, commit to a direction, and move — knowing you might be wrong, knowing you'll have to adjust, knowing that the adjustment itself is the strategy. Boyd's framework gives that willingness a structure and a name. It makes "act under uncertainty" feel like a method rather than a gamble. That psychological shift is worth more than the diagram.
The failure mode I see most often isn't speed fetishism or locked orientation — it's what I'd call "loop collapse." Teams start with all four phases. Within a few cycles, Orient and Decide merge into a single instinctive reaction. Within a few more, Observe narrows to only the metrics that confirm the previous decision. The loop doesn't stop spinning — it just stops learning. You can diagnose this by asking a simple question after any OODA cycle: "What did we learn in the Orient phase that surprised us?" If the answer is "nothing" for three consecutive cycles, the loop has collapsed into a reflex. Reflexes are fast. They are also brittle. The moment the environment shifts in a way your reflex doesn't cover, you're exposed.
The highest-leverage modification is to separate the Orient phase into two distinct sub-steps: destruction and creation. Boyd actually described this in his briefings but most popularisations omit it. Destruction means deliberately breaking apart your existing mental model — questioning assumptions, identifying what's changed since your last orientation, surfacing contradictions in your data. Creation means synthesising a new model from the pieces. Most teams do creation without destruction: they update their existing model incrementally, adding new data to old assumptions. That's patching, not reorienting. The discipline of destroying before creating — even for five minutes per cycle — is what keeps the Orient phase honest. It's uncomfortable. It's supposed to be. Comfort in the Orient phase is the first sign that your loop has stopped working.
Section 9
Top Resources
01
Boyd: The Fighter Pilot Who Changed the Art of War — Robert Coram (2002)
BookThe definitive biography and the closest thing to a primary source on Boyd's thinking, since Boyd himself never wrote a book. Coram traces the development of the OODA Loop from energy-manoeuvrability theory through Boyd's Pentagon briefings. Chapters 18–24 cover the mature OODA framework in detail, including the Orient phase's internal structure that most business writers omit. Essential reading — not just for the framework, but for understanding why Boyd believed orientation was the decisive phase.
02
Certain to Win: The Strategy of John Boyd, Applied to Business — Chet Richards (2004)
BookRichards studied directly under Boyd and this is the best translation of OODA into business strategy. He maps Boyd's concepts to competitive dynamics, organisational agility, and product development cycles with specificity that generic strategy books lack. The chapter on "operating inside the competitor's OODA Loop" is the clearest explanation of tempo-based competition available outside military literature.
03
Book
Ries's Build-Measure-Learn loop is the OODA Loop adapted for product development, whether he intended the parallel or not. Build maps to Act, Measure to Observe, Learn to Orient. Reading the two frameworks together reveals what each one emphasises that the other underweights: Ries is stronger on validated learning and minimum viable products; Boyd is stronger on competitive tempo and the psychology of orientation. Together they form a complete operating system for startups in contested markets.
04
Book
Grove's concept of "strategic inflection points" is an Orient-phase manual. His account of Intel's decision to exit the memory business and bet on microprocessors is one of the best documented examples of an organisation breaking its existing mental model (destruction) and building a new one (creation) under competitive pressure. Grove never cites Boyd, but the structural parallel is exact — and the book is a masterclass in what happens when orientation succeeds and when it fails.
05
Book
The scientific foundation for why the Orient phase is both the most important and the most failure-prone part of the loop. Kahneman's System 1 (fast, automatic, pattern-matching) is what takes over when the loop collapses into reflex. His System 2 (slow, deliberate, analytical) is what the Orient phase demands. The chapters on anchoring, availability, and the planning fallacy explain the specific cognitive biases that lock orientation — and why structural interventions (like Boyd's destruction-before-creation discipline) are necessary to override them.
Decision Tools Library — Browse by phase
Why this matters next
mental modelsSecond-Order Thinking
Cynefin Framework applied the Second-Order Thinking mental model
mental modelsConfirmation Bias
Cynefin Framework applied the Confirmation Bias mental model
mental modelsFirst Principles Thinking
Cynefin Framework applied the First Principles Thinking mental model
mental modelsLeverage
Cynefin Framework applied the Leverage mental model
mental modelsOODA Loop
Cynefin Framework applied the OODA Loop mental model
mental modelsNarrative
Cynefin Framework applied the Narrative mental model
Continue exploring
More like this, in your inbox
I send a newsletter every week — free, no spam, unsubscribe anytime.