How a solo attorney tracks billable hours on a Mac without a timer
· 7 min read
The problem with timers in a practice is not that they are hard to use. It is that the day does not cooperate. A client calls during drafting. Opposing counsel emails something that needs an answer now. You start the timer for one matter and finish an hour later having touched three.
The result is familiar: at the end of the week you reconstruct Tuesday from memory, and the reconstruction is conservative, because guessing high on a client’s bill is not a thing careful people do. Under-recording is the standard failure mode, and it is expensive.
A different approach: stop trying to capture the allocation in the moment, and capture the raw shape of the day instead. Allocate afterwards, from a record rather than from memory.
What a record of the day gives you
Automatic tracking on a Mac notes which application was in front and for how long. At the end of the day you get a list: the word processor for two hours forty, mail for one hour ten, the browser for fifty minutes, the PDF reader for half an hour.
That is not a bill. It says nothing about which matter any of it belonged to. What it does give you is the outer boundary and the shape: how much time was actually at the desk, and how it divided between kinds of work.
Reconstructing from that is a different exercise from reconstructing from nothing. You are no longer asking “how long did I spend on the Henderson matter” with only memory to go on. You are asking “the record says two hours forty in drafting on Tuesday; how did that divide between Henderson and Whitfield”, which is a question memory can actually answer, because you remember the sequence even when you cannot remember the durations.
Durations are what people estimate badly. Sequence is what they remember well. Playing to that is the whole method.
The daily step, which takes four minutes
At the end of the day, with the record in front of you, write the matter allocation while it is fresh.
Take each block in the record and assign it. Two hours forty of drafting becomes ninety minutes Henderson, seventy Whitfield. Mail becomes twenty minutes across four matters plus fifty on practice administration that nobody is paying for.
Do it the same day. The record keeps for weeks; your memory of what you were drafting does not survive the weekend.
Four minutes a day is roughly twenty minutes a week, and it produces a defensible allocation rather than a plausible one.
The number that changes how you price
Once you have a few weeks of this, an uncomfortable figure emerges: total hours worked against total hours billed.
The gap is administration, business development, reading that was not chargeable to anyone, the call that turned out to be a prospective client who did not instruct, and the twenty minutes lost to each interruption that nobody could bill for.
Most people in solo practice are surprised by this number the first time, and it is the number that matters for pricing. An effective hourly rate is what you actually earn divided by the hours you actually worked, not your nominal rate. If the nominal rate is comfortable and the effective rate is not, the answer is rarely to work more hours. The time audit to do before raising your rates walks through that calculation.
What this does not do
Worth being direct about, because the alternative is a tool that promises more than it delivers.
An application-level record does not know which matter you were working on. It sees a word processor, not a document. It cannot produce a matter-by-matter breakdown by itself; the allocation is a human step, done daily, and if you will not do it the record stays raw.
It also does not integrate with practice management software or generate an invoice. It produces a record of the day; turning that into a bill happens wherever your bills are produced.
If you need matter codes captured at the moment of work, with the software knowing the client, then dedicated practice management software with a timer is the right tool and this is not a substitute for it. The case for the approach here is for practitioners who have tried that and found that the timer does not survive contact with a real day.
Where Punchcard fits
Punchcard sits in the menu bar, notices which application is in front by name only, and at the closing time you set prints a receipt of the day: one line per application with its time, and a day total.
It asks for no macOS permissions. It never reads window titles, document names, URLs or keystrokes, which for a practice matters more than it does for most work: a tool that recorded document titles would be recording client names and matter descriptions, and that is a confidentiality question you do not want to have. There is no account, no cloud and no network code at all, so the record does not leave the machine. Why a time tracker should never need screen recording covers what other tools ask for and why.
What it gives you is the raw shape of the day, reliably, without you remembering to start anything. The allocation to matters is yours to do, and it is the four-minute step above.
Practical notes
Set the closing time realistically. The receipt prints then, and a closing time you routinely blow past is still useful, because the overrun shows as an overrun. Setting a closing time covers picking one.
Keep the receipts. A folder of them is a record you can go back to if a bill is queried months later. Proving your hours when a client disputes an invoice covers what that is worth.
Court days and travel will not be captured. Time away from the Mac does not appear. Note those separately; the record covers desk work, which is most but not all of it.
Questions
Is a record of application usage sufficient support for a bill? It is corroboration, not a substitute for contemporaneous records. Your matter notes remain the primary record. What the receipt does is make those notes accurate rather than reconstructed.
What about client confidentiality? An application-level record contains no client information, because it never collects any. That is the point of the boundary. A tool that captured window titles would capture matter names, and that is a very different thing to have on a laptop.
I work on paper as much as on screen. Then the record covers part of your day, and you add the rest by hand. It is still better than reconstructing everything.
Does this help with fixed-fee work? More than with hourly, arguably. Fixed-fee pricing is only as good as your estimate of how long the work takes, and most people have never measured. A few months of records turns pricing from instinct into arithmetic.