Guide
A work log is a dated, running record of what you worked on, what you finished, what you decided and what is still open. A plain text file is enough. To skip end-of-day recall, Timing suits billable Mac work, ActivityWatch is free and cross-platform, Toggl fits teams on timers, and Dayflow writes a readable daily timeline from your screen for $0 on Mac.
A work log is a dated, running record of the work you did: what you worked on, what you finished, what you decided and what is still open. It is written for your future self first, and for anyone else only when you choose to share it.
People get it wrong in two directions. Some treat it as a timesheet and record only hours, which tells you how long something took but not what happened. Others treat it as a diary and write paragraphs about how the day felt, which is pleasant to write and nearly useless when you need to find the date a decision was made.
Disclosure: Dayflow, one of the tools covered near the end of this guide, is our product.
A work log is a chronological file of short entries, usually one per day, that describe your work in enough detail to be searchable later. Four features separate it from ordinary notes:
It goes by other names: work diary, daily log, engineering notebook, lab notebook in research settings. A “brag document” is a close cousin that keeps only accomplishments. The underlying shape is the same: a date, then what happened.
The value shows up later. A work log answers questions memory handles badly. When did we agree to drop that feature? What did I actually do the week the deadline slipped? What should go in my self-review? Without a log, you rebuild those answers from email, chat history and guesswork.
These four get mixed up because they overlap. Each answers a different question.
| Main question | Unit | Usual reader | Typical rhythm | |
|---|---|---|---|---|
| Work log | What did I do, and what changed? | Short dated entries | You, sometimes a manager | Daily |
| Timesheet | How many hours went where? | Hours per project or client | Payroll, clients, finance | Daily or weekly |
| Journal | How did the day go for me? | Free-form reflection | You | Whenever |
| Status report | What does the team need to know? | Progress, risks, blockers | Manager, team, stakeholders | Weekly or per standup |
The useful way to see it: the work log is raw material for the other three. If you note rough times, you can build a timesheet from it. A weekly status report is mostly your last five entries, edited down. Reflection can sit in a short section at the bottom of an entry without taking over.
The reverse does not work. Hours on a timesheet will not tell you what was decided. A status report written for an audience leaves out the dead ends you will want to remember. A journal entry about a frustrating day rarely says which ticket caused it.
Six fields cover almost everything. You do not need all six every day.
Optional extras that earn their place for some people:
What to leave out: passwords, keys, other people’s personal data, and client details you are not allowed to store outside approved systems. That matters more if your log syncs to a cloud service.
Knowing the reader decides how much detail to write. There are usually five.
You, tomorrow morning. The most frequent reader. Yesterday’s “next step” line is what gets you moving without spending the first half hour reconstructing context.
You, at review time. Performance reviews, promotion packets, portfolio updates and contract renewals all ask what you accomplished over months. A log turns that from an exercise in recall into an exercise in selection. If you regularly struggle with this, see our guide on how to remember what you worked on.
Your manager or team. They rarely want the raw log. They want the edited version: a standup line or a weekly summary. A log makes those fast to write, and our piece on automatic standup updates covers turning daily records into team updates.
Clients. When a client asks what an invoice covered, a dated log of outcomes answers better than a column of hours.
Future you in a disagreement. When someone asks when a decision was made or who approved a change, a dated entry written at the time carries weight that memory does not.
The hardest of these is future you without context. Write for that reader and the others are covered.
There is no best format, only the one you will keep. Four common options:
One file, one dated heading per day, newest at the top or bottom. It opens instantly, works in any editor, searches with a keystroke and will still be readable in twenty years. This is the right default for most people.
A template gives you the same headings every day, which helps if you tend to skip fields like decisions or blockers. Our daily work log template uses the six fields above. Pair it with the weekly status report template and the Friday roll-up becomes a copy-and-trim job.
Good when you need to total hours or filter by project or client. Weaker for decisions and context, because long text in cells is awkward to read and write. Some people keep a spreadsheet for hours and a text log for everything else.
Software records what you did on the computer, and you edit that into an entry instead of writing from memory. This solves the most common failure, which is forgetting to write anything. The tradeoff is that software sees activity, not intent, so you still add decisions and next steps by hand. Our explainer on what an automatic time tracker is covers how this kind of capture works.
Call it Capture, Close, Carry. It takes a few minutes spread across the day plus a short weekly step.
Step 3 is where most of the value comes from. Rough capture without a close leaves you with notes you cannot read in a month. A close without capture forces you to rely on memory, which is exactly what the log is supposed to replace.
This is an illustration, not a real person or real data. Imagine Maya, a freelance product designer working for two clients.
Her captured fragments during the day:
9:10 onboarding flow v3, client A
10:40 call w/ A PM, they want to cut step 3
11:30 email B re: icon set delay
13:00 onboarding v3 revisions
15:20 stuck on empty states, need copy from A
16:30 B icon set, 6 of 20 done
Her closed entry at 5 pm:
## 2026-10-05
Client A - onboarding flow
- Revised v3 based on PM call. Removed step 3 (account type picker).
- Decision: account type now defaults to Personal and is changeable in settings.
Reason: PM says most sign-ups never change it, and step 3 was the main drop-off point in their own review.
- Blocked: empty-state copy. Asked PM at 3:20 pm.
Client B - icon set
- 6 of 20 icons done. Told B the set moves to Friday (emailed 11:30).
Next: send A the v3 link with the step-3 change called out.
A month later, when client A asks why the account type picker disappeared, the answer and its reason are in one dated entry. When Maya invoices client B, she can show what was delivered and when the delay was communicated. Neither fact would survive on memory alone.
Logging activity instead of outcomes. “Worked on onboarding” is activity. “Removed step 3 from onboarding” is an outcome. Activity entries feel complete when written and say almost nothing later.
Writing it on Friday from memory. Weekly reconstruction misses the small things and flattens everything into the two or three most memorable tasks. If your sense of where the week went often disagrees with what you produced, the where does my week go calculator gives a rough picture to compare against.
Too many fields. A template with twelve headings gets abandoned within a fortnight. Start with date, outcome and next step, then add fields only when you notice you need them.
Writing for an audience. Once a log is written to impress a manager, the dead ends, mistakes and uncomfortable decisions disappear, and those are often what you most need later. Keep the log honest and edit when you share.
Storing it where you cannot search it. Paper notebooks are pleasant and nearly impossible to search. Logs split across five apps get the same result. Pick one searchable place.
Mixing in sensitive data. Client secrets, credentials and colleagues’ personal details do not belong in a log, especially one that syncs off your machine.
You do not need software to keep a work log. Tools help mostly with the capture step, by recording a raw history of your day that you then turn into entries. If you want the broader comparison, our guide to the best Mac time trackers goes further. The details below come from comparing each vendor’s own documentation as of October 4, 2026.
| Tool | What it does for this | Price |
|---|---|---|
| Timing | Records time across apps, websites and documents on Mac automatically; AI activity summaries and AI-suggested time entries | Professional from $9 per month billed annually or $11 billed monthly; no free tier; 30-day trial |
| ActivityWatch | Records active app, window title and AFK status; browser extensions add tab title and URL; data stays on your device | Free, open source (MPL-2.0) |
| Toggl (formerly Toggl Track) | Timers, timesheets and calendar integrations, plus background tracking and a timeline in the desktop app | Free for up to 3 users; Starter $9 per user per month billed annually or $12 billed monthly |
| Dayflow | Captures one screen frame every 10 seconds and has AI write the day into a readable timeline; exports a date range as Markdown | $0, MIT licensed; optional Pro plan at $20 a month |
Timing is a Mac-only automatic tracker that stores tracked times locally unless you opt into sync. Its Expert plan is $11 per month billed annually or $14 billed monthly, and the Connect plan for teams is $16 per user per month billed annually or $20 billed monthly. It is built around billable-hour records, with billing status tracking and report rounding. See our Timing comparison.
ActivityWatch is free and open source, runs on Windows, Mac, Linux and Android, and keeps all data on your device. It lists no built-in AI summaries, so the raw record of apps and window titles is yours to turn into entries. See our ActivityWatch comparison.
Toggl (formerly Toggl Track) is a cloud service centered on timers and manual entry, available on Web, Mac, Windows, iOS, Android and browser extensions. Its Premium plan is $16 per user per month billed annually or $24 billed monthly, and Enterprise pricing requires contacting the vendor. The vendor states it does not take screenshots or record keystrokes. See our Toggl for Mac comparison.
Dayflow is our product. It fits if you are on a Mac running macOS 14 or later and want a written timeline of your day to edit into a work log, with recordings stored on your Mac and the choice of local AI, your own cloud provider or Dayflow Pro for analysis. It does not fit if you need invoices, payroll approvals, team reports or website blocking, and its AI summaries can miss or misread work, so check entries before you share them.
A work log is a dated record of outcomes, decisions and next steps, kept mainly for your future self. It is not a timesheet, a journal or a status report, though it feeds all three. Start with one file and three fields: date, what changed, what is next. Close each day in five minutes and roll up once a week. If the habit keeps failing because you forget to write anything down, let a tool capture the raw day and spend those five minutes editing instead of remembering.
At minimum: the date, what you worked on, what changed as a result, any decisions and the reason behind them, open blockers, and the next step. Times and links are useful extras. Keep each entry short enough to write in about five minutes, and focus on outcomes rather than activity, since 'shipped the fix' is far more useful later than 'worked on bugs'.
A timesheet answers how many hours went to which project or client, usually for payroll or billing. A work log answers what you did and what changed, in words. You can often build a timesheet from a detailed work log if you note times, but a timesheet alone rarely tells you why something took as long as it did or what was decided.
Most useful entries are five to fifteen lines. Long enough to name the work, the outcome, any decision and the next step; short enough that you will actually write one every day. If entries regularly run past a page, you are probably writing a journal or a report. Move reflection into a separate section and keep the log itself factual and searchable.
Usually not the raw log. Treat the log as private working notes and share edited summaries, such as a standup update or a weekly status report, built from it. That keeps the log honest, because you are not writing for an audience, and gives your manager the shorter, decision-focused version they actually need. Some teams do share logs openly, which is fine if everyone agrees.
The best format is the one you will keep daily. A single plain text or Markdown file with one dated heading per day works for most people because it is fast, searchable and portable. A template helps if you tend to forget fields. Spreadsheets suit people who need to total hours. Automatic capture suits people who forget to write anything at all.
Partly. Automatic trackers can record what was on screen or which apps and windows were active, and some use AI to summarize it. Dayflow captures one frame every 10 seconds and has AI write a readable timeline, exportable as Markdown. Timing lists AI activity summaries. You still need to add decisions and next steps yourself, and AI summaries can miss or misread work, so check them.
Yes. ActivityWatch is free and open source under MPL-2.0, stores data on your device, and records active apps and window titles. Dayflow is $0 and MIT licensed, with an optional Pro plan at $20 a month, and needs macOS 14 or later. A plain Markdown file in any text editor is also free and works well if you write entries by hand.
Not Dayflow, which has no invoicing, billable rates or team reports. On Mac, Timing tracks billing status, offers report rounding and, depending on plan, integrates with GrandTotal for invoicing. Toggl offers billable rates and team reports on paid plans and runs on Web, Mac, Windows, iOS and Android. Pick Timing for a solo Mac setup and Toggl for a mixed-platform team.
Leave out passwords, API keys, personal data about other people, and client details you are not allowed to store outside approved systems. Be especially careful if your log syncs to a cloud service. Also skip venting about colleagues. A work log may be read later, by you or someone else, and it should hold up as a factual record of the work.
Keep it at least until your next performance review or contract renewal, since that is when it pays off most. Many people keep years of logs because plain text takes almost no space and old entries answer questions like when a decision was made. If your employer or clients have record retention rules, follow those instead.
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