Eisenhower matrix with a workload check

Two questions split the list. Urgent means there is a deadline and it comes from outside: the reply is expected today, the release is booked, a customer has asked. Important means the task moves a goal you are accountable for. They are different axes, and crossing them gives four sets with four different actions.

Important and urgent

Yours, today. This quadrant is supposed to be small: if half the list is in it, the matrix is not showing you priorities, it is showing you that the work has become firefighting.

Important, not urgent

The only quadrant that changes next month: specs, churn analysis, agreeing a process, learning something. Every item here needs a date in the calendar — without one it waits for a deadline and arrives in the first quadrant instead.

Not important, urgent

Other people’s deadlines and routine that someone else can carry. Hand it over with a definition of done or it comes back. And if there is nobody to hand it to — which happens more often than management books admit — the quadrant becomes a second “do” list, and its hours belong with the first.

Not important, not urgent

Delete. A task that has sat here for six months will not become important; it is only taking up room in your head. Move these to a separate someday list and do not open it weekly.

A plain layout says nothing about whether the week is survivable: nine tasks in the first quadrant look exactly like three. That is why each task here carries an estimate and the week carries a number of hours.

Hours available is not 40. Take out meetings, on-call, the inbox and the things that simply happen. Most people are left with 15 to 25, and that is the number worth comparing a list against.

The warning fires at 55%: once the first quadrant takes more than half the time you have, there is physically no room for “important, not urgent”. A week later those items hit their deadlines and move into the first quadrant, and round it goes. The threshold is a working rule rather than a finding — tune it to yourself, but keep some threshold.

With no estimates the tool says plainly that it is counting tasks only. “About two hours” beats no estimate at all: a wrong estimate is visible immediately, while a missing one hides the overload.

It is a tool for choosing between tasks already in front of you, and it has edges. A few cases where it honestly will not help:

• The flow of work is not yours to control. In support, in an agency, on an on-call rota, nearly everything is urgent because the deadlines are set outside. The matrix will show an overflowing first quadrant — that is a diagnosis of the process, not of your planning. It is fixed by response-time agreements and priority rules, not by a better layout.
• “Important” is decided by who asked. If importance tracks the seniority of the requester, the matrix turns into an org chart. You need a goal to check against first; without one, all four quadrants are filled by instinct.
• More than twenty-five tasks. At that point it is not a matrix, it is a list in four columns. Split it by week or group it by project, and use the matrix on one week at a time.
• Nobody to delegate to. For a solo founder or the only specialist in a team, the third quadrant is fiction. Those tasks still have to be done, so their hours belong in the count.
• You need to compare the size of the effect. The matrix cannot say which of two important tasks returns more: it has no reach, no impact, no effort. That is what RICE and ICE scoring is for, and for choosing between product features, the Kano model survey analyzer.
• One task, but a big one. “Rewrite billing” does not belong in any quadrant until it is broken into steps. The matrix works on tasks measured in hours, not on projects measured in quarters.

One more observation that does not invalidate the method but explains part of its failures: people reliably pick the urgent task over the important one even when the important one pays better. Zhu, Yang and Hsee called it the mere urgency effect. That is why the first quadrant grows so easily — urgency is felt more strongly than value.

Any planning system lives until the first day nobody opens it. What helps:

• Sort weekly, not daily. Ten minutes at the start of the week: carry over what is unfinished, cross out what has gone stale, give the second quadrant dates.
• A date instead of a quadrant. An item in “important, not urgent” counts as handled once it has time in the calendar. Skip that step and the quadrant becomes a warehouse of good intentions.
• Say the drops out loud. “Not doing this” is a decision too, and worth recording. Otherwise the same task returns in a month, as urgent.
• Export to where the work lives. Markdown into a wiki or a ticket description, CSV into a spreadsheet or a tracker where rows become tasks. Print is for the version three people argue over in a room.
• One list, not four. The matrix is a view of a list, not a separate system. If tasks live in a tracker, sort them here before planning and take them straight back.

If what needs sorting is not your work but the steps your users have to take inside a product, a different instrument applies: an onboarding checklist shows a person the next step where they already are, and asks them for no matrix of their own.

1. Covey S. The 7 Habits of Highly Effective People, 1989, ch. 3 — dividing work by importance and urgency, and the central claim about quadrant two. Covey credited the older importance/urgency distinction rather than Eisenhower himself.
2. Eisenhower D. Address at the Second Assembly of the World Council of Churches, 1954 — “I have two kinds of problems, the urgent and the important”, the line the matrix is named after. Eisenhower never drew the matrix.
3. Zhu M., Yang Y., Hsee C. The Mere Urgency Effect, Journal of Consumer Research, 2018 — people choose urgent tasks over more rewarding non-urgent ones even when the difference in payoff is obvious.
4. Covey S., Merrill A. R., Merrill R. First Things First, 1994 — planning by the week, and the practice of booking time for quadrant two in advance.