Comparing by fit
Two ways of thinking, side by side. There is no winner here — read across each row and choose the one that fits your situation.
Decision Making
Observe, orient, decide, act — and win by cycling faster than the situation changes.
By 6days
Operations
Separate urgent from important, and notice how much of your week serves neither.
By 6days
When to use — OODA Loop
Use it in genuinely competitive, fast-changing situations — an incident, a live negotiation, a competitor's surprise move, a crisis. It suits environments where information is incomplete by nature and waiting for completeness means losing.
When to use — Eisenhower Matrix
Use it when you are busy but not progressing, when a team is permanently reactive, or during a periodic review of where a role's time actually goes. It works well as a recurring habit rather than a one-off, and it is a useful device for a manager and report to look at a workload together without it becoming personal.
When not to use — OODA Loop
Avoid it for decisions that are expensive to reverse and slow-moving: a factory site or a pension scheme does not want tempo, it wants analysis. It is widely misread as 'decide fast', which drops the orient step and produces speed without judgement. It also has little to say where there is no adversary and no clock.
When not to use — Eisenhower Matrix
Avoid it where you have little real autonomy — telling someone to delegate work they cannot delegate is just a way of blaming them for their constraints. It handles individual tasks better than long collaborative efforts, and it has nothing to say about work that is important to someone else and not to you, which is most of what makes a job hard. It also assumes you can tell importance from urgency, which is exactly the skill people struggling with this tend to lack.
In fast-moving situations the quality of a decision decays with time. A team gathers information, builds consensus, and commits — and by the time it acts, the situation has moved and the decision addresses a world that no longer exists. Meanwhile a less careful competitor has already acted three times, learned from each, and is operating against a picture of reality that is simply more current than yours.
Urgency is loud and importance is quiet. Work that matters — the strategy, the fix that prevents next quarter's incident, the conversation with the struggling team member — has no deadline, so it loses every day to whatever is shouting. Weeks pass full of activity and empty of progress, and the diagnosis is always 'no time', when the truth is that the important work never once won a scheduling contest against the merely urgent.
The OODA loop describes decision-making as a continuous cycle: observe what is happening, orient by making sense of it against your models and experience, decide on a course, act, then observe the results and go again. Its central claim is comparative — in a contest, whoever cycles faster and more accurately accumulates an advantage, because their opponent is perpetually responding to a stale picture. Orientation is the pivot and the part most often skipped: it is where prior beliefs distort what the observations mean, and a fast loop feeding a broken orientation just produces confident errors at speed.
The matrix sorts work on two independent axes: important (contributes to outcomes you care about) and urgent (demands attention now). The four quadrants get different treatment. Important and urgent: do it, but treat its existence as a signal, since most crises are the residue of neglected important work. Important, not urgent: schedule it, because this is where nearly all real progress lives and it will never demand attention on its own. Urgent, not important: delegate it, or resist it — it is usually someone else's priority wearing your calendar. Neither: stop. The insight is that urgent and important are genuinely independent, and most people's calendars silently assume they are the same axis.
Gather what is actually happening — raw signals, not the summary someone assembled to support a position. Most organisations observe through reporting layers that have already discarded the anomalies, which are precisely the observations that matter.
Interpret the observations against your models, and deliberately ask what frame you are imposing and what it might be hiding. This is the step that determines everything downstream and the one under time pressure that gets skipped. A fast wrong orientation is worse than a slow right one.
Commit to a course knowing you do not have everything. Waiting for certainty is itself a decision — usually the worst one available, because it hands the tempo to whoever is willing to move.
Execute in a way that generates information. A well-chosen action resolves uncertainty as well as making progress, which means the next observation step starts from a better place than the last.
An ordered process with 4 phases.
Gather what is actually happening — raw signals, unfiltered.
Make sense of it, and challenge the frame you are imposing.
Commit to a course while information is still incomplete.
Worked example — OODA Loop
During a partial outage, one team convenes a call to establish full root cause before acting. Another observes that errors are concentrated in one region, orients on a recent config change to that region's load balancer, decides to roll it back despite not having proven causation, and acts within nine minutes. The rollback resolves it — and had it not, the negative result would itself have eliminated the leading hypothesis. The first team is still on the call, assembling a complete picture of a situation that the second team has already ended.
Worked example — Eisenhower Matrix
An engineering manager audits a week. Urgent and important: a production incident, four hours. Urgent, not important: eleven meetings where they were informed rather than needed, nine hours. Important, not urgent: hiring, the architecture decision blocking two teams, and a struggling report's development conversation — zero hours, as every week. Neither: two standing meetings inherited from a reorganisation eighteen months ago that nobody has cancelled. Declining six meetings and booking two protected blocks moves nine hours into the quadrant that was empty. The incident, it later turns out, was caused by the architecture decision that never got made.
The loop was developed by John Boyd, a United States Air Force colonel and military strategist, from the 1970s onward, initially to explain air combat outcomes and later generalised into a broader theory of competition and adaptation. Boyd published little formally, working mainly through briefings, so the model circulates through interpreters — and popular versions often flatten the orientation stage he considered central.
The urgent-versus-important distinction is popularly attributed to Dwight D. Eisenhower, who quoted a formulation of it in a 1954 address and credited it to an unnamed university president. It was later developed into a four-quadrant tool by Stephen Covey and others in the time-management literature of the 1980s and 1990s. Eisenhower did not devise the matrix that bears his name.
Feed results straight back into observation, and treat cycle time as a metric in its own right. The advantage does not come from any single pass; it comes from the rate.
List the actual demands — tasks, meetings, requests, the recurring things you have never questioned. Work from your real calendar and inbox rather than from memory, which reliably omits the routine items that consume the most hours.
Importance is meaningless in the abstract, so name the outcomes you are accountable for and judge each item against them. Anything that contributes to none of them is not important, however comfortable or familiar it is.
Ask what genuinely happens, and to whom, if this waits a week. Most 'urgent' items have a soft consequence and a loud sender. Rating this independently of importance is the whole mechanism; collapsing the two axes reproduces the problem you came to solve.
Do the important-and-urgent, delegate or decline the urgent-but-not-important, delete the rest — and then actually put the important-not-urgent work in the calendar as a booked commitment. If it stays on a list it will lose again next week, exactly as it lost last week.
Execute so the action itself teaches you something, then loop.
A two-by-two matrix. The horizontal axis runs from Urgent on the left to Not urgent on the right; the vertical axis runs from Important at the top to Not important at the bottom. Top-left: Do now — Crises and hard deadlines, Pressing problems you must handle. Top-right: Schedule — Planning and strategy, Relationships and prevention. Bottom-left: Delegate — Interruptions and some meetings, Requests that are someone else's priority. Bottom-right: Delete — Busywork and time-wasters, Habitual activities that add nothing.