How to Track Writing Progress: 5 Methods Compared

There are five ways to do this: a tally on paper, a bullet journal, a spreadsheet, a web service with an account, an app on your device. Four of them are free at entry level. And the one randomised trial that put two of these media against each other over 24 months concluded that sticking with it matters more than which one you pick.

So the short answer is: choose the one you will still be using in three months. Which one that is depends less on the tool than on where you write. What follows is what each method can do, what it costs, and where it falls apart.

The five methods side by side

What you put in those columns, whether words, characters, pages or time, is a separate decision. It belongs in What a writing tracker is and who it helps and it should be settled before the tool question. Pick the wrong unit and every one of the five methods will show you a red day.

How this comparison was made

What was checked on 8 September 2026 is what the vendors publish themselves: pricing pages, knowledge bases, the TrackBear repository through the GitHub API. Anything that could not be read there appears here as a gap rather than as a number. A long-run test in which the same people keep all five methods in parallel would be a different thing, and no such test exists for this subject. What does exist is a randomised trial covering two of the five media. It comes later, and it does not say what any of the five camps advertise.

1. A tally on paper

A sheet, a date column, a number column. That is the whole system. No account, no battery, no vendor that might close. Writing the entry takes as long as writing a four-digit number.

The weakness is not the paper. It is the analysis. Ninety handwritten numbers do not become a curve, a weekday average or a pattern without typing them somewhere else. If you want to know after a quarter whether Sundays are worth planning for, you first have to move the data.

Paper is still the sensible choice for the first four weeks, which is not what most comparisons recommend. During those weeks you do not yet know how big a typical session is for you or which days actually produce anything. Choosing a tool before those numbers exist means configuring it from guesses. Four weeks of paper cost nothing and hand you the two figures every other tool asks for.

2. Bullet journal

The Bullet Journal is a method rather than a template. Ryder Carroll defines it through Rapid Logging, the Index, the Future Log, the Monthly Log, the Daily Log, Collections and a monthly Migration. The official guide lists the equipment as “A notebook and something to write with”.

For writing projects the useful unit is the Collection: one spread per project, daily numbers going in. What it adds over a loose tally is the Index, so you can still find March in November. What it adds over software is the Migration, because once a month it forces a question that apps tend to swallow: am I still carrying this project, or is it dead?

The price is time. A bullet journal wants maintenance, and the monthly migration costs a quarter of an hour in which nothing gets written. If you already keep one for daily life, you are paying that anyway. If you start one purely for word counts, you now run two systems.

3. Spreadsheet

The most recommended route, and it can cost nothing. LibreOffice Calc is open source and free, Google Sheets runs in a browser with an account. A spreadsheet averages, charts and sums by month. Five minutes of setup gets you exactly the columns you want.

Its limit shows up where the entry is made. Somebody who writes 400 words on a train in the evening rarely opens a spreadsheet to log them, and the reminder to log at all has to come from somewhere else, because a spreadsheet does not ask.

There is a further point that comparisons almost never raise. A spreadsheet you built yourself is only as good as its formulas. A range that stops one row short in an average will go unnoticed for months, because the result still looks plausible. A finished application cannot make that mistake, and in exchange you cannot inspect its arithmetic.

4. Web service with an account

This group is crowded, so here are three examples that were actually checked.

TrackBear counts words, chapters, pages, scenes, lines and time, runs hosted, and can be self-hosted with Docker. The source sits openly on GitHub. This is where a closer look pays off: the repository contains no licence file, and the GitHub API reports no licence for the project. Publicly readable source is not the same thing as open source software. Anyone self-hosting it has been granted no formal permission to do so, however detailed the setup instructions in the repository are.

4thewords wraps counting in a role-playing game where words defeat monsters. The vendor states its free tier as 25 files, 2 projects, one month of file version history and limited access to battles. That is an honestly documented limit, and it is one that a second book breaks.

Pacemaker turns a deadline, a target and your unavailable days into a daily workload and keeps adjusting it. Its knowledge base names Free, Lite and Pro tiers. The prices are not on a page that can be read without signing in, so they are unverified here and are not quoted.

5. App on your device

An app measures where the writing happens, does the arithmetic, and does the reminding. It puts widgets on a home screen, runs several projects at once, and turns timestamps into a heatmap. Above all it shortens the distance between finishing a session and logging it to a few seconds, and by the evidence below that distance matters more than anything else on this page.

In return it wants a device, usually an account, and trust. Three questions belong before the install: does the data sit on the device or with the vendor, is there an export and what does it cost, and does the app keep working if the vendor stops?

What the research actually says

Two pieces of work carry weight here, and they do not say the same thing.

The meta-analysis by Harkin and colleagues pools 138 experimental studies with 19,951 participants and finds a medium effect of progress monitoring on goal attainment, d = 0.40 with a confidence interval of 0.32 to 0.48 (Psychological Bulletin 142(2), 198–229, 2016). The effect is larger when progress is physically recorded and reported to someone. That is the case for measuring at all, whatever the medium.

On the medium itself there is exactly one trial that randomised people onto two of these five methods. Burke and colleagues followed 210 overweight adults for 24 months across three conditions: a paper diary, a handheld device, and the same device with daily tailored feedback. Only the third group lost a significant amount of weight (p = 0.02). The authors’ conclusion: “adherence to self-monitoring is more important than the method used” (American Journal of Preventive Medicine 43(1), 20–26, 2012).

Three limits belong with that sentence. The trial was about diet and weight, not writing. A sample of 210 is a decent study, not proof. And the group that won was not the digital one, it was the digital one with feedback. Moving from paper to a screen on its own did nothing.

For choosing a tool, that lands somewhere uncomfortable. The question is not paper against app. It is which method gives you an answer the same day about the number you just entered.

What three years cost

Paper and spreadsheets are quickly dealt with: materials, or nothing at all. The paid models are where a sum is worth doing, and almost nobody does it.

Those are WriteZone’s prices, the tracker behind this blog, checked on 8 September 2026. Other vendors differ in level, rarely in ratio. Paying monthly costs three times the yearly price over three years, and more than three times the one-off. If you know you will still be writing in a year, the decision is already made.

Before anyone pays, though, the free tier decides, and its limit is almost always the number of projects: two at 4thewords, one active project at WriteZone. A person with a blog, a book and a report is out immediately. A person with one manuscript often stays free forever. The cost question is rarely about features. It is about how many things you write at once.

A second sum, with no money in it: three writing days a week gives you 156 entries a year. After twelve months a weekday average rests on 52 values per weekday, after three months on 13. Below roughly ten values per weekday the number is an anecdote. That is why a tracker tells you nothing in week two.

The point most comparisons skip

On 31 March 2025 NaNoWriMo dissolved. The organisation cited longstanding financial problems, and the website has been gone since. For more than two decades hundreds of thousands of people had logged their word counts there, and for many of them that was the only place their writing years existed (TechCrunch, 1 April 2025).

The lesson is not to avoid web services. It is one question that belongs before you sign up rather than after: how do I get my numbers back out, and in what format? A CSV file will still open in any spreadsheet in ten years. A chart on somebody else’s website is unreadable the moment the website goes.

Paper has an advantage here that no software matches. It is offline, it depends on no vendor, and it is still legible in a hundred years. It also burns.

Running two methods at once

Plenty of people keep two rather than one, and that is not indecision. The entry is made where the writing happens; the analysis happens where the arithmetic is. A slip of paper in a notebook plus a weekly transfer into a spreadsheet is a working combination, as long as the transfer has a fixed slot.

It has a predictable failure point. Every extra step between session and number is a place the chain can break, and a weekly transfer tends to break in the busy week, which is exactly the week whose data would have been interesting. So hang the transfer on something that happens anyway: the end-of-week review, Sunday evening, a slot already in the calendar.

The other workable pairing runs the other way: app for the daily number, spreadsheet for the annual review. That one only works with an export, which puts you back at the question from the previous section.

When each method is the wrong one

1. Paper, if you want analysis. The numbers are there, but the answer to “which days actually produce” costs you an evening of typing.

2. Bullet journal, if you do not already keep one. A notebook purely for word counts is a second system next to the one you have.

3. Spreadsheet, if you write away from your desk. The entry happens where the text happens, or it does not happen.

4. Web service, if your connection is unreliable or you would rather not attach your manuscript to an account.

5. App, if you have exactly one project finishing in four weeks. A sheet of paper covers that.

What to do with this

Use paper for four weeks, whichever method you want afterwards. At the end of it you will know how big a typical session is and which days produce anything. Those two numbers let you configure any other tool sensibly, and the writing goal calculator will tell you whether your deadline fits them. What daily targets are realistic in the first place is covered in How many words a day.

More analysis and the research behind it lives under Data & Studies, conversions and metrics under Word Goals & Metrics.

Share