The deep work season: four weeks around one deliverable

Deep work fails most often not inside the block but around it. People read about long, undistracted sessions, try to fit them into an ordinary week, and end up with a scattering of good mornings and no finished thing. A deep work season fixes the frame instead of the session: four weeks, one deliverable, a fixed number of blocks a day, and an end date that arrives whether or not you feel ready.

One deliverable, not a direction

The season is built around a thing that will exist on the last day. Not "get better at writing" or "make progress on the app", which are directions, but an object you could hand to someone: a 12,000-word report in a file, a checkout page that takes real payments on staging, a grant application submitted on the 30th.

The test is simple. On day 28, could a stranger tell whether you did it? If the answer depends on how you feel that day, the sentence is still a direction. Rewrite it until it can fail.

One deliverable also means the others wait. You may have three projects that matter. A season holds one of them in the deep hours and lets the rest live in the shallow ones, or in the next season. Splitting the blocks between two deliverables is the most common way to finish neither.

Count blocks, not hours

Hours are a poor unit for deep work because they include everything: the time spent opening the file, answering a message, rereading yesterday, getting coffee. Four hours "on the project" can contain forty minutes of actual thinking.

A block is a different unit. It has a start time, a length (ninety minutes is a good default), a single task named in advance, and a rule: nothing else enters. The phone is in another room. The browser holds only what the task needs. When the block ends, it ends, even if it is going well.

At night you don't write "worked 4 hours". You write "two blocks: drafted section 3, fixed the payment webhook". That record is harder to inflate and much easier to learn from. After a week you know roughly what a block produces, and the plan for the remaining weeks stops being a guess.

Two blocks is the ceiling

The temptation, especially in week one, is to schedule three or four blocks a day. It looks serious on paper. By Thursday it produces a crash, a lost day, and the quiet conclusion that you are not cut out for this.

Two blocks a day, every day, for four weeks is around fifty-six blocks. That is a great deal of deep work, more than most people do in a quarter. The ceiling exists because the quality of the second block depends on how you treated the first, and the quality of tomorrow depends on how you treated today.

If you have a free day, lengthen the blocks a little. Do not add a third. If you have a day with almost no time, one block is a full day. Zero blocks is a lost day, and a lost day is just that: not a reason to do double tomorrow, which fails, and not a broken streak, because there is nothing to break.

A shape for the four weeks

The four weeks are not identical, and knowing their shape in advance makes them easier to hold. There is a fuller week-by-week version in this 4-week plan; the short version for a deep work season is this.

• Week 1: one block a day. Install the schedule. The goal is seven days in which the block happened at its hour, not a large output.
• Week 2: two blocks a day. The structure now holds, so the load rises. This is where the deliverable usually turns out bigger than it looked. Narrow the work, not the date.
• Week 3: the same as week 2. Novelty is gone and the end is not yet visible. Expect sudden urges to reorganise the plan or research something adjacent. Do the same blocks at the same hours.
• Week 4: finishing only. No new sections, no new features. What is half done gets completed or cut, and the deliverable leaves the building before the last day.

What to do with the rest of the day

Two blocks take three hours. The other waking hours are not wasted; they are what makes the blocks possible. A season that ignores them burns out by week three.

Shallow work still exists. Email, meetings, admin and errands go in their own window, ideally after the first block and never before it. A useful rule: nothing that someone else wrote enters your head until block one is done. No inbox, no news, no messages.

The body needs a daily slot. Thirty minutes of walking counts, without headphones if you can manage it; a lot of the thinking that blocks depend on happens there. Sleep is not the reserve you raid when the deliverable is behind. It is what the deliverable runs on.

And leave real slack. One half day a week with no blocks and no plan. The schedule survives because it includes rest on purpose instead of taking it by collapse.

A sample day

Times are illustrative; adjust them to your life and then stop adjusting.

• 07:00 Up. No phone for the first hour.
• 07:45 to 09:15 Block one. The hardest part of the deliverable, named the night before.
• 09:15 to 13:00 Shallow work: messages, meetings, admin.
• 13:00 to 14:00 Lunch and a walk.
• 14:00 to 15:30 Block two. The rougher half: editing, tests, data, the mechanical parts.
• 15:30 to 18:00 Everything else the day needs.
• 21:30 Ten minutes of review, then screens off.

If you have a full-time job, the same two blocks move to the margins: one before work, one in the evening, and on many days only the first. One real block a day for four weeks still adds up to a deliverable.

The nightly count

Each evening, before bed, write three lines. How many blocks happened, and what each one moved. What you avoided, named plainly: "spent block two reorganising the outline instead of writing section 4". And the first task of tomorrow's first block, specific enough that you can start without deciding anything.

That last line is the hinge of the whole season. Most blocks that fail, fail in their first five minutes, while you work out what to do. Deciding it the night before removes that window. If you want the full version of this habit, the nightly review covers it in detail.

When the season ends

On the last day, compare the sentence from day one with what exists. Done or not done, and by how much. Count the blocks you actually ran against the fifty-six you planned; the gap usually tells you more about your schedule than about your discipline.

Then stop. Take at least a week without blocks before you consider another season. The gap is not a reward. It is what keeps the next four weeks a season instead of a permanent state you quietly abandon in month two.

Pick the deliverable, write the sentence, and put the first block in tomorrow's calendar.

If you want a four-week season with the blocks scheduled and someone asking each night what they produced, La Orden is in free beta.