How to manage a long writing project

Short pieces are finished by effort. Long ones are finished by systems, because no amount of motivation survives twelve months. The failure mode is not dramatic — almost nobody abandons a thesis in a decisive moment. They stop for a fortnight, the re-entry cost grows, and the fortnight becomes a year.

Everything below is aimed at that one failure.

Make the project visible

Break it into pieces small enough to finish in one sitting. Not "chapter 3" but "the section on sampling limitations, about 800 words". A task you can complete today produces momentum; a task that takes three weeks produces avoidance.

Keep the whole plan on one page or one board with every piece and its state: not started, drafted, revised, done. Seeing thirty items with eleven marked done is what carries you through month seven.

Track cumulative words, not daily ones. A graph that only goes up is much better for morale than a daily target you will sometimes miss.

Protect the routine, not the mood

Same time, most days. The specific time matters less than that it is decided in advance and defended. Writing when you feel like it works for about three weeks.

A small daily minimum beats a large intermittent one. 500 words a day is a book a year. Two thousand words on the good days and nothing otherwise is not, because the good days are rarer than you think and the recovery after a gap costs more than the gap.

Never stop at a natural break. Stop mid-section with a note about what comes next. Re-entry is the expensive part of every session and this makes it nearly free.

Keep a five-minute version. On bad days, do five minutes. It is not about the output; it is about not breaking the chain, because the chain is the only thing holding the project.

Separate the modes

Drafting, researching, editing and formatting are different activities and doing them together is slow and demoralising.

Never research mid-sentence. Write `[CHECK: 2019 figure]` and continue. Do the checks in a batch later.

Never edit while drafting. Keep a "later" file for every problem you notice. The file grows and none of it stops you.

Do formatting last, once. Rewriting citations as you go is wasted work if the section gets cut.

The middle

Every long project has a stretch where the beginning is unsatisfying, the end is invisible, and the work feels pointless. This is a stage, not a verdict, and knowing it is coming reduces its power considerably.

What works: switch to a different section for a week; reread your own best material rather than someone else's; return to the one-page plan and tick things off; talk it through with someone who will ask questions rather than give advice.

What does not: starting a new project, restructuring the whole thing, or rereading from the top.

Version control and safety

Date every file: `thesis-ch3-2026-09-05.docx`. Never overwrite a major draft — disk space is free and you will want to recover a deleted paragraph.

Back up in two places, one of them off your machine, automatically. Every long-project writer has one story about a lost draft and it is always avoidable.

Keep a single decisions log: what you cut and why, what you decided about scope, what you promised a supervisor or editor. In month nine you will not remember, and you will otherwise relitigate settled questions.

Working with a supervisor or editor

Send work on a schedule, even when it is not ready. Long silences are how projects drift, and a supervisor cannot help with work they have not seen.

Send with a specific question. "Does the chapter 3 argument hold?" gets a useful reply; "any thoughts?" gets a delayed and vague one.

Agree deadlines for their feedback too, and plan around the lag.

Common mistakes

Waiting for a clear month. Rewriting the opening repeatedly — it should be written near the end, when you know what it introduces. Reading more instead of writing, which feels productive and is the commonest form of avoidance. No backups. Perfecting section 1 while section 9 does not exist. Hiding from a supervisor after a bad month, which makes the next contact harder still.

Practical tips

• Put the deadline and the word count on the wall.
• End every session by writing the first line of the next one.
• Schedule a weekly half-hour review: what moved, what is stuck, what is next.
• Take one full day off a week, deliberately. Unplanned collapse is worse than planned rest.
• Tell people the date you will finish. It is uncomfortable and it works.