Guide
Thesis months usually disappear to a goal too big to start, reading that feels like progress, and long stretches without feedback. Write a small bad draft every day against a kitchen timer, keep a three-line paper log, and send your supervisor something every two weeks. On a Mac, Dayflow can show where the hours actually went. A paper log works anywhere.
It is March. The thesis was supposed to start properly in October. You have a folder of PDFs highlighted in three colors, a document called draft_v3_FINAL that is mostly headings, and a vague sense that the last few months were busy. When your supervisor asks how it is going, you say “mostly reading,” and that is true.
Then a day goes like this. You open the draft, reread the introduction, fix two sentences, check one citation, follow it into another paper, answer some email, and at 5 pm the word count has moved by forty. Nothing went wrong exactly. The day just never turned into thesis, and neither did the one before it.
This article covers why a thesis is so good at swallowing time, seven things to try that mostly need nothing more than paper, a calendar and a kitchen timer, and one place where an automatic record of your day can help. That last part mentions Dayflow, which is our product and the publisher of this blog.
A thesis has almost none of the structure that keeps other work moving. There is one deadline, and it is far away. For most of the year nobody is waiting on anything specific from you this week, so nothing feels urgent until everything is.
The task is also too big to see. “Write a thesis” cannot be started on a Tuesday afternoon. When the next step is unclear, the brain reaches for something that is clear: reading one more paper, reformatting references, tidying the folder. These feel like work, and some of them are, but they rarely produce pages.
Reading deserves its own mention. It is the most respectable form of avoidance in academia. There is always one more source, and reading never exposes you to judgment the way a draft paragraph does. Many people find that the closer the writing gets to “real,” the more urgent the reading becomes.
Then there is feedback, or the lack of it. In a job, a wrong turn gets noticed in days. In a thesis, you can spend six weeks on an approach your supervisor would have questioned in five minutes, simply because you did not show anyone.
Finally, unstructured days blur together. Without a record, it is hard to tell whether last week had ten hours of writing or two. Memory tends to keep the feeling of being busy and drop the details, which is why so many people end a month unsure where their days went.
None of this means you are lazy or bad at research. It means the shape of the work is unusual, and it helps to add some shape back.
Break the next piece down until the first step is obvious and slightly boring. A good test: could you start it in the next two minutes without deciding anything? If not, it is still too big.
Each evening, write tomorrow’s first task on a sticky note and put it on your laptop. One task, phrased as a physical action: “Open results.docx and write three sentences describing Figure 4.”
Illustration: Instead of “Literature review,” the sticky note says “List the four papers that disagree with my argument, one line each.” That takes twenty minutes and gives you the skeleton of a section.
After a week: Count how many mornings you started the sticky-note task within ten minutes of sitting down. If it is most of them, keep going. If not, your tasks are probably still too large, so cut them in half again. If starting itself feels physically hard, our guide to task paralysis and how to get started has more ideas.
Set a timer for 25 to 45 minutes and write in the thesis document, nothing else. No reading, no checking citations, no fixing earlier sections. Put a placeholder like [CITE] or [CHECK] where you would normally stop and go look something up.
Do this before you open email or a single PDF. Once reading starts, writing rarely gets its turn.
The draft is supposed to be bad. Its job is to exist, so that later you have something to improve. Editing a weak paragraph is much easier than writing a perfect one from nothing.
Illustration: Someone writing a methods chapter sets a kitchen timer for 30 minutes at 9 am, writes 400 rough words with six [CITE] markers, and spends the afternoon filling in those markers. The reading now has a purpose.
After a week: Look at the document. Is there more text than there was? Ignore quality for now. If the daily block keeps getting skipped, make it shorter, not longer. Fifteen minutes that happen beat an hour that doesn’t.
Ending a session at a natural break, like the end of a section, makes the next start harder, because tomorrow you face a blank heading. Stop instead in the middle of a sentence or paragraph you know how to finish, and write a one-line note to your future self.
Illustration: The last line in the document at 6 pm reads: “The second reason the model underperforms is that the training data… [NEXT: finish this, then compare with Table 3].” The next morning, finishing that sentence takes thirty seconds, and you are already writing.
After a week: Notice how long it takes to get going each morning. If the gap between sitting down and typing the first new word has shrunk, this is working. Some people find it uncomfortable to leave things unfinished. Try it for five days before deciding.
At the end of each working day, write three lines in a notebook:
That’s it. Two minutes. The point is not accountability to anyone else. It is having a record you can look back on when the week feels like a blur, and catching the “stuck” items before they quietly cost a month. If you prefer a structure, a daily work log template gives you the same idea in a ready format, and our guide to keeping a work journal goes deeper.
Illustration: On Thursday you look back and notice “stuck: not sure the survey data supports claim 2” appears on Monday, Tuesday and Wednesday. That is not a writing problem. It is a question for your supervisor, and now you know to ask it.
After a week: Read all seven entries in one sitting. If you can see a pattern you did not know about, such as afternoons that never produce writing or one problem that keeps reappearing, the log is earning its two minutes.
Do not wait until a chapter is “ready.” Send your supervisor something written every two weeks or so: a page of rough argument, a revised outline, a list of three questions. Ask what rhythm suits them, and put the send dates in your calendar like deadlines.
Small, frequent submissions do two things. They create the near-term deadlines a thesis lacks, and they catch wrong turns while they are still cheap.
If your supervisor is hard to reach, a peer works too. A friend in your program, a writing group, or someone who just sits with you while you both work can supply the structure. Our piece on body doubling and how to try it covers that last idea.
Illustration: Two students agree to meet at the library every Tuesday and Thursday at 10 am, spend the first five minutes saying what they will write, write for an hour, and spend five minutes saying what they did.
After a week: Did you produce something because someone was expecting it? If the external date made you write, keep the dates and make them regular.
Take your submission date and work backward: final proofread, supervisor’s last read, revisions, each chapter draft. Write the dates in a paper calendar or on a wall. Then add buffer, because almost everything in a thesis takes longer than it looks.
It helps to know how far off your sense of time runs. The time estimation test is a short exercise comparing your felt sense of time with a clock. It is not a diagnostic, just a nudge. If you consistently underestimate, our article on why we underestimate how long things take explains the usual reasons.
Illustration: You guess the discussion chapter will take three weeks. Before committing, you check how long the much shorter results section actually took, which was four weeks, and you plan six for the discussion instead.
After a week: Compare what you planned for this week with what got done. If the gap is large every week, the plan is the problem, not you. Stretch the timeline now rather than in month eleven.
Pick a fixed slot, such as Friday at 4 pm. Read your log, look at the calendar, and answer three questions: What moved? What didn’t, and why? What is the one thing that matters most next week?
Keep it short and kind. This is a planning meeting, not a trial. If you want a more structured look at where the hours go, the time audit template walks through a week in more detail.
Illustration: Friday’s review shows that every morning with a 9 am meeting produced no writing at all. Next week you move the writing block to 7:30, before the meeting.
After a week: The first review itself is the test. If it changed even one decision about next week, it is worth twenty minutes.
The paper log in step 4 works when you fill it in. On the days that matter most, the ones that dissolved, people often can’t. Rebuilding a day from memory tends to produce the version you wish had happened.
Dayflow is a free Mac app (macOS 14 or later, MIT licensed) that records your screen at one frame every 10 seconds and has AI write the day into a readable timeline. There is no timer to start or forget. At the end of a “writing afternoon,” you can see that it was forty minutes in the draft, an hour following citations and a long stretch in email, then plan tomorrow around that instead of around guilt.
Full recordings and the timeline are stored on your Mac. With Local AI, through Ollama or LM Studio, screenshot analysis stays on-device too. If you choose your own cloud provider or the optional Pro plan at $20 a month, selected screenshots are processed off your Mac, as explained in local and cloud processing. You can export a date range as Markdown from Settings → Export, which is handy before a supervisor meeting. The AI summaries can miss or misread things, so check them before sharing.
Dayflow is our product, and this article is published on its blog. The product details above come from Dayflow 2.2.0’s documentation as reviewed on 2026-09-06.
It is Mac only. If you write on Windows, a Chromebook or an iPad, a paper notebook does this job better than any app we could recommend here.
It only sees the screen. Reading printed papers, annotating by hand, lab or field work, and thinking on a walk all happen off-screen, and a lot of thesis work looks like that. If most of your work is away from the laptop, the three-line paper log wins.
It is a record, not a coach or a blocker. Dayflow does not block websites and will not stop you mid-scroll. If in-the-moment distraction is the main problem, a website blocker is the better tool. If what you lack is accountability, a writing group or a study partner beats any software.
It is also not a planner. For the backward plan in step 6, a paper calendar is enough.
If your thesis involves sensitive participant data, check your ethics approval and data rules before running any screen recorder. Use Local AI or leave it off on those days. If you want to see other approaches, our roundup of Mac time trackers compares the options.
Writing a thesis is stressful, and it can overlap with poor sleep, persistent anxiety, low mood, burnout, or attention difficulties such as ADHD. This article is not medical advice. If you notice that stress has stopped being about the thesis and started affecting your sleep, eating, mood or ability to function, a doctor or licensed clinician is the right person for assessment and treatment. Many universities have a counseling service, and your supervisor or program office can usually point you to it. For the writing itself, a university writing center is often a good place to start.
A thesis rarely gets lost in one dramatic failure. It leaks away through vague days, comfortable reading and long stretches without anyone seeing the work. The fix is mostly unglamorous: a small, concrete task each morning, a short daily writing block against a kitchen timer, three lines in a notebook each evening, and something sent to your supervisor every couple of weeks.
Start with one of these tomorrow, not all seven. Pick the one that matches where your time actually goes. If you’re on a Mac and genuinely can’t tell where the hours went, an automatic record can fill in what memory leaves out.
There is no correct number, and long days are not the goal. What tends to help is one protected block of actual writing most days, often 45 to 90 minutes, with reading, data work and admin around it. A steady short block you keep beats a heroic eight-hour day you dread and then avoid for a week. Adjust the length once you see what you really sustain.
Reading feels productive, has no blank page, and never exposes you to judgment. Writing does all three. Many people keep reading because one more paper feels like preparation, while a draft paragraph feels like a test. Try capping reading to a set time, keeping a 'to read later' list, and writing a rough paragraph about each source before you open the next one.
Write to find out what you think, not to record what you already know. Start with a messy paragraph that begins 'What I think is going on here is...' and keep going for 25 minutes without fixing anything. You can also talk the argument out loud to a friend or voice memo and transcribe it. Clear thinking often arrives during drafting, not before it.
It depends on your program and your supervisor's style, so ask them directly. Whatever the meeting schedule, sending something written every two weeks or so keeps the work real. It can be a page, an outline or a list of questions. Regular small submissions catch wrong turns early, which is where a lot of lost months come from.
Skip the full reckoning and start with the next two weeks. Write down what exists now, pick one chapter or section to move forward, and set one small daily writing block. Tell your supervisor plainly where things stand and what you will send them next. A revised plan built from today is far more useful than guilt about the past.
A light record helps more than strict tracking. Three lines on paper each evening, covering what you did, what is next and where you are stuck, is enough to show patterns within a week. If you forget to write things down, an automatic record such as Dayflow on a Mac can fill gaps. Avoid systems so detailed that maintaining them becomes another way to not write.
No. Dayflow records what was on your screen and turns it into a timeline afterward. It does not block websites or apps, and it will not pull you back to your draft in the moment. If in-the-moment distraction is your main problem, a website blocker is the better tool. Dayflow is more useful for seeing, at the end of a day or week, where the time actually went.
No. Dayflow is Mac only and needs macOS 14 or later. If you write on Windows, a Chromebook or an iPad, a paper notebook beats it: a few lines each evening about what you did and where you got stuck gives you most of the same awareness. The habit of looking back at your week matters more than the tool that captures it.
Check your ethics approval and your institution's data rules first, since interview transcripts or participant records may have strict handling requirements. With Dayflow, recordings stay on your Mac, and the Local AI mode keeps screenshot analysis on-device. Using your own cloud provider or Dayflow Pro sends selected screenshots off your Mac. If participant data cannot leave your machine, use Local AI or do not run it on those days.
Dayflow is $0 and open source. Full recordings stay on your Mac, with local AI available for on-device analysis.
Download for MacOS macOS 14+ · MIT licensed · no account required