Do you need a project picker in a time tracker? Probably not

· 6 min read

If you bill by the project, you need a project picker; there is no way around it. If you track time to understand your own day, the picker is the main reason you will stop tracking within a month, because every switch demands a decision and the log ends up half wrong. Most people tracking for themselves do better with an app-level record and a two-minute review at the end of the day.

What a project picker really asks of you

A project picker looks harmless on the screenshot: a dropdown, a few names, a start button. In use, it asks for a decision every time your attention moves. Reply to a client email: is that “Client A” or “Admin”? Open the design file to check a measurement for a different job: switch the project, or let it ride? Take a call that covers two contracts: split it afterward, or guess?

Each decision is quick. The problem is the count. A normal day has dozens of these moments, and a tracker that needs a choice at each one turns into a chore you start skipping by Wednesday. Once you skip a few, the log is wrong, and a wrong log is worse than none, because you will make decisions on it anyway.

The second thing the picker asks is that you know what you are about to do before you do it. Real work is not like that. You open the editor intending to fix one bug and come out two hours later having refactored a module. The project was the same, but the task changed, and the picker never knew.

Who genuinely needs one

Be honest about this part, because the picker is not a bad idea, it is a specialized one.

You need projects if:

  • You invoice clients by the hour and they expect hours broken down by job.
  • You work for an agency that bills your time to different accounts.
  • You are on a grant or contract that requires time allocation by category.
  • A manager asks you to report hours per initiative.

In every one of those cases, someone other than you needs the split, and the split is the product. A picker, tags, or a manual timer are the price of producing it. Trackers built for this are built around the picker for good reason, and a tracker without one will not do the job.

What the log looks like without projects

Drop the picker and what remains is the thing your Mac already knows: which app was in front and for how long. The record for a day looks like this:

  • Editor, 4 h 10 min
  • Browser, 1 h 40 min
  • Mail, 50 min
  • Calendar and video calls, 1 h 20 min
  • Messaging, 35 min
  • Terminal, 30 min

No decisions were made to produce that list. It was collected by noticing the front app and adding up the minutes. It is also honest in a way a project log rarely is, because there was no moment where you could round up or forget to stop the clock.

What it lacks is intent. The list says the browser was in front for an hour and forty minutes; it does not say whether that was research or drift. That is the trade: you give up the “why” in exchange for a record that is complete every single day.

Getting project-level answers from an app-level record

For most personal questions, the app-level record is enough, and you can get surprisingly close to project answers with a short review. Here is the method.

  1. Pick the day’s closing time and look at the record once, after it has ended. Do not look during the day.
  2. For each line, ask one question: does this app belong to one kind of work? The editor probably means building. Mail and messaging probably mean coordination. Calls mean meetings.
  3. Add the lines into two or three buckets that matter to you, such as making, talking, and everything else. Write the three numbers in a note or on the receipt.
  4. For the one ambiguous line (usually the browser), estimate the split from memory. You will be within a quarter of an hour, which is better than most project logs.
  5. After a week, read the three numbers across seven days. That is the trend you actually wanted from the picker.

This works because the questions people ask themselves are coarse. “Am I spending the day in meetings?” “Is the editor in front for more than three hours?” “Where did Friday go?” None of those need a dropdown. They need a complete record and a moment to read it. Freelancers have a version of the same method in time tracking for freelancers without projects or tags.

Where Punchcard fits, and where it does not

Punchcard is built on exactly this idea: there is no project picker because there is nothing to pick. It notices which app is in front during your day (app names only, never window titles, document names, or URLs) and, at the closing time you set, prints a paper receipt itemized by app with a time per line, a day total, and a stamp. Sundays bring a week roll, and the end of the month brings a month roll. You never start or stop anything, and there are no tags.

It also asks for zero macOS permissions: no Accessibility, no Screen Recording, no Full Disk Access. That is possible precisely because it is not trying to read window titles to guess your project.

And this is where it stops. Punchcard has no per-project or per-client tracking, no billable rates, no invoicing, no manual timers, and no team features. If you are in the “someone else needs the split” group from the section above, it will not produce what they need, and you should use a tracker built around the picker. If your question is “where did my day go,” the receipt answers it without ever asking you a question back. The trade-off between the two styles is laid out in timer-based time tracking vs automatic.

Questions

Can I get both: automatic tracking and projects?

Some trackers try, usually by reading window titles or URLs and matching them against rules you write. It works when your projects map cleanly onto documents and sites, and it requires the tracker to read a lot more about you than the app name. Decide whether that trade is worth it for the split you need.

I only have one client. Do I still need projects?

Almost certainly not. With one client, the split you need is usually “billable versus not,” and an app-level record plus a one-line note at the end of the day gets you there. If you bill by the hour, check how to track billable hours without forgetting to start the clock for the honest version of that workflow.

What if I want to track a side project separately from my job?

If they happen in different apps or at different hours, the app record already separates them. If they happen in the same editor at the same desk, a project picker is the only tool that will split them, and you will have to accept the daily decisions that come with it.