Time Management

Why One Running Timer Beats Three Open Clocks

Time tracking for freelancers usually dies in the same place: you are “on” three clients in your head, so you start three clocks, or you leave one clock running while you answer another Slack, or you run no clock and promise you will reconstruct the afternoon later. Reconstruction is where invoices get fictional.

The tempting product answer is simultaneous timers — one per client, like browser tabs. It feels like honesty about context switching. It is actually a delay of the decision: which task was this ten minutes for? AP does not pay “I was juggling.” They pay a named unit of work.

What three open clocks are actually doing

Parallel clocks usually encode anxiety, not accuracy. You are afraid the unused clock will forget a client. So you keep them all “warm,” like pots on a stove, and you stir whichever one the latest message belongs to.

• You never stop anything. Stopping would mean admitting the other two were idle theater.
• Overlapping intervals appear on the invoice. Two clients charged for 2:15–2:40 is a conversation you will lose.
• The task name stays vague because the clock was started from a client, not from a unit of work.
• Friday you average. “Split it 40/40/20” is not measurement. It is diplomacy with your future self.

Consultants who work retainers feel this harder: the week looks busy, the clocks look busy, and the report still cannot answer which project ate Thursday. If you already know a calendar is not a timesheet, treat extra clocks the same way — they are a picture of intention, not of time.

One timer forces the only question that bills

A single running timer asks, every time you switch: what is the leaf work right now? In a client → project → task (and subtask) world, time belongs to a task. If you cannot name the task, you are not ready to start a clock. You are ready to send a message, or to take a note, or to decide this is overhead.

The second line is slightly more annoying in the moment. It is dramatically less annoying when you generate a report. You get a sequence, not a Venn diagram.

Context switching is still real — log it as switches, not as overlays

You will bounce. Designers bounce from a Slack “quick look” to a Figma file. Developers bounce from a build failure to a second repo. Strategists bounce from a doc to a call. The honest timesheet is a series of short entries, not one mashed interval.

If the bounce was 4 minutes and you do not want 4-minute line items, that is a rounding and grouping conversation with yourself — not an excuse to keep two clocks alive. Rounding, in tools that support it, is a display rule on stored originals. It is not a license to overlap.

A day that actually switches clients

Here is a freelance Tuesday that does not need three clocks:

1. Start on Homepage mockups for Contoso when you sit down.
2. A Northwind ping is genuinely urgent: stop, start the Northwind task (or log a short manual entry after if it was two minutes of triage).
3. Back to Contoso: start the mockups task again. Two entries for one task in one day is fine. That is what happened.
4. Lunch is not a client. Stop.
5. Afternoon workshop: one task, one clock. If the workshop spans two clients, that is two tasks — which means a stop in the middle, even if it feels fussy.
6. End of day: History. Anything you did without a clock gets a manual entry while the memory is still warm.

That is more start/stop clicks than a fantasy of omnipresent clocks. It is fewer lies. It also matches how you already work on a phone between sites: you are not actually coding for two companies in the same second. You are alternating.

Do not confuse the focus clock with the billable clock

Pomodoro (25/5) is a focus metronome. It can sit on a dashboard as a widget and still have nothing to do with who you invoice. We wrote about that split in Your Pomodoro Clock Is Not Your Invoice. The running timer is the billable clock. One of those. If you treat Pomodoro cycles as invoice lines, you will either underbill the messy middle or overbill the breaks.

What to stop tracking so one timer stays tolerable

People abandon time trackers because every context switch feels like paperwork. Shrink the catalog so the clock is easy to start:

• Track leaf work, not vibes. “Contoso” is a client, not a task. “Homepage mockups” is a task.
• Park overhead on overhead tasks. Sales calls, invoicing, tool wrangling — if you care about real hourly rate, give them a home. Do not glue them onto a client clock “so it isn’t lost.”
• Lock finished phases. A closed task cannot collect more time. That is how revision rounds stop infecting “final” files.
• Search, then play. If starting a timer takes a tour of folders, you will not stop the previous one. Global search that can start a task is the difference between a habit and a chore. See dashboard widgets if you want the timer next to a to-do list you already intended to work.

Onboarding in a tool with a real hierarchy is still the same four moves: add a client, add a project, add a task, start the task. If you skip to “start something,” you will name it later, and later is when three clocks are born.

A one-week experiment

If you are currently a three-clock person, try this without changing your rates:

1. Monday–Wednesday: force one running timer. When you switch clients, stop then start. No overlapping clocks even if the tool would allow it.
2. Thursday: generate one report for the week so far. Pivot by task. Notice whether the story matches the days you remember.
3. Friday: invoice from that report if the week is closed — or at least compare it to last Friday’s reconstructed split.

Most people do not miss the extra clocks. They miss the feeling of safety. Safety is History plus a report, not a choir of running stopwatches. Getting started is short on purpose: the hierarchy is the product, and the timer is just how minutes enter it.