Meetings versus making: measuring the split on your own Mac
· 6 min read
To measure the meetings-versus-making split, add up the time your calling app was in front over a normal week, then add up the time your making apps (editor, design tool, writing app) were in front, and compare the two totals. The calendar is not a reliable source for either number, so the measurement has to come from what was actually on screen. A week is the minimum; a single day is too noisy to mean anything.
Why the calendar is the wrong place to look
Most people who feel over-scheduled go to their calendar first. It looks authoritative: every meeting has a start, an end, and a duration. But the calendar records invitations, not attendance, and it misses almost everything around the meeting itself.
A thirty-minute call starts three minutes late, runs eight minutes over, and is followed by ten minutes of writing up what was decided. The calendar says thirty minutes. The day says fifty. Multiply that by four meetings and you have found more than an hour the calendar never mentions.
It works the other way too. Calendar blocks labeled “focus time” get eaten by a quick question, and the calendar still claims you spent two hours making. The only honest source is the screen: which app was in front, and for how long.
Decide what counts as making before you measure
If you decide afterward, you will make the number look the way you want it to. So settle this first.
Making is the work that produces the thing you are paid for. For a developer that is the editor and the terminal. For a designer it is the design tool. For a writer it is the writing app. For someone who lives in spreadsheets, it is the spreadsheet.
Meetings are simpler: the calling app, plus whatever you use for screen sharing if it is a separate app.
Everything else (email, chat, browser, calendar, file management) is a third bucket. Do not try to split it into “meeting-adjacent” and “making-adjacent”. That distinction will not survive contact with a real week. Three buckets is enough: meetings, making, and the rest.
One caveat: if your meetings happen inside a browser tab, they will land in the browser’s total, not a calling app’s total. See the section on limits below.
A one-week measurement, step by step
- Pick a normal week. Not the week before a launch, not a week with two days off. If nothing is normal, pick next week and accept it.
- Write down your guess now, before measuring. Something like “meetings 30 percent, making 50 percent, rest 20 percent”. Put it somewhere you will find it again.
- Get a per-app record of each day. You can do this by hand (a note with a timestamp every time you switch), with Screen Time in System Settings (it reports per app, though it is built around limiting use rather than reviewing work), or with a menu bar tracker that records the front app automatically.
- At the end of each day, copy three numbers into a note: minutes in the calling app, minutes in your making apps combined, and the day total.
- After five workdays, add each column up. Divide meetings by the total and making by the total. Those two percentages are your split.
- Compare with the guess from step 2. The gap is usually the most useful thing you learn.
Do not skip step 2. A measurement with no prior guess produces a number you shrug at. A measurement that contradicts your guess produces a change.
Reading the ratio honestly
There is no correct ratio. A manager at 60 percent meetings is doing the job. An individual contributor at 60 percent meetings probably is not, and also cannot fix it alone.
What the ratio is good for is spotting three patterns.
First, the shape across the week. If Tuesday and Thursday are meeting days and the making happens Monday, Wednesday, and Friday, the average hides a week that actually works. If meetings are spread evenly across every day, no day gets a long stretch, and the making total will be low even when the meeting total looks moderate.
Second, the third bucket. People who measure expecting to find too many meetings often find the meetings are fine and the rest (chat, email, the browser) is what ate the making time. That is a different fix, and Why you feel busy all day and still got nothing done goes into it.
Third, the making number on its own. Two hours of making in an eight-hour day is common and often a shock. It is worth knowing even if you cannot change the meeting load, because it changes how much you should promise.
Doing it with a receipt instead of a spreadsheet
The manual method works, and for one week it is fine. The part that fails is step 4: copying numbers at the end of the day, every day, for five days. Most people manage Monday and Tuesday.
Punchcard exists for this. It sits in the menu bar and notices which app is in front, by app name only. At the closing time you set, it prints a receipt of the day: one line per app with the time, and a day total at the bottom. You read the calling app line, you read the editor line, and you have your two numbers without having recorded anything yourself. On Sunday it prints a week roll, which is the five-day sum from step 5 done for you.
It asks for no macOS permissions at all, so it never sees window titles, meeting names, or what was on screen. That is also the reason for the first limit in the next section.
It does not have a “meetings” category or any way to tag apps into buckets. You do the bucketing by reading the lines, which for three buckets takes about ten seconds. If you want per-project or per-client splits, it is not the tool; it only answers “where did the day go”.
What this measurement cannot tell you
Any per-app record, whether kept by hand or automatically, has the same blind spots. Be clear about them so you do not over-read the result.
Meetings in a browser tab look like browser time. If your calls run in a browser, the meeting bucket is undercounted and the rest bucket is overcounted. The fix is to use the desktop app for calls during the measurement week.
It cannot tell whether a call was useful. A weekly one-to-one and a status meeting that could have been a message both show up as thirty minutes.
It cannot tell whether the editor time was making. Staring at a broken build is editor time. So is reading code you wrote last month. The number is an upper bound on making, not a measure of output.
And it only sees the Mac. Meetings taken on a phone while walking, or in a room with no laptop open, are invisible. If those are a big part of your week, note them by hand and add them in.
Questions
What ratio should I aim for?
There is not one. Use the measurement to check whether your week matches your role, and to see which days get a long uninterrupted stretch. A 40 percent meeting load spread evenly across five days is worse for making than 50 percent packed into two.
Does one week really tell me anything?
One week tells you the shape. It does not tell you the trend. If the first week surprises you, measure a second one a month later before changing anything large. How many hours do you really work? A one-week measurement is the companion to this post for the day-total side.
Can Punchcard split my calls by project or client?
No. It records app names and time, nothing else. It has no projects, tags, or rates. If you need billable splits, you need a tool with a project picker, and you will need to keep the picker honest yourself.
My meetings are in the browser. Is the measurement useless?
Not useless, but the meeting bucket will read low. Either switch to the desktop calling app for the week, or treat browser time as “meetings plus browsing” and accept a rougher number.