Spotting the person who never stops: hours without a break

Most people buy a time tracker to find out who is doing too little. I built one, and the pattern I did not expect is the opposite: the person who never stops. If you track work hours without breaks — meaning you look at the gaps in the day, not just the total — the quiet overworker shows up in the data long before they show up in a resignation email or a doctor's visit.

This matters to me personally. I have been in front of a screen for about twenty years. For a long stretch of that I worked in unbroken blocks: sit down at 9, look up at 7, eat at the desk. Nothing hurt, so nothing seemed wrong. Then it did hurt, and the part that hurt did not go back to normal. Wrists, neck, eyes. You do not get a warning the day before. You get a decade of fine, and then a permanent bill.

A small remote team makes this invisible. In an office you notice the person still at their desk when you leave. Remotely, you notice nothing. They are "online". That is all you see.

What "no break" actually looks like in tracked data

A total of 9 hours tells you almost nothing. Two people can both log 9 hours and have completely different days:

• Person A: 9:00–12:30, gap, 13:30–18:00. Two blocks, one real lunch.
• Person B: 8:40–18:20 with the longest gap being 6 minutes. Nine hours and forty minutes of continuous activity.

Person B is the one to worry about, and the daily total barely distinguishes them. So stop reading totals and read these four things instead.

1. Longest continuous block

Take the activity timeline for the day and find the longest stretch with no gap bigger than, say, 10 minutes. That single number is the most useful health signal in the whole dataset. My own threshold: anything over 4 hours gets my attention, over 5 hours I say something.

2. Number of real gaps

Count gaps of 15 minutes or more. A normal day has two or three — a lunch, a coffee, something. A day with zero is not a productive day; it is a day where someone skipped lunch at their keyboard.

3. Span from first to last activity

First keystroke 8:10, last one 20:40 — that is a 12.5-hour span even if the tracked time is only 8 hours. The gaps in the middle were probably childcare or dinner, not rest. The workday is still eating the whole day.

4. The pattern across the week

One intense Tuesday is a deadline. Four weeks of Tuesdays is how someone works. Look at a month, not a day. The screenshots and app data help here in a narrow way: if the long block is one app, it is usually a grind task (spreadsheets, support queue, QA). Those are the tasks people forget to stand up during. I wrote about reading that kind of data without turning it into judgment in how to read screenshot data without micromanaging.

How to use this without creeping anyone out

Here is the trap. The moment you say "I noticed you worked 11 hours straight," your employee hears you are watching me, not you care about me. Intent does not travel; the surveillance reading always arrives first. So the delivery matters more than the data.

What has worked for me:

Tell them this is one of the reasons you track. Say it when you introduce tracking, not after you catch something. Something like: "I look at the hours for billing and for workload. If someone is logging 10-hour days I will bring it up, because that is my problem to fix, not yours." That reframing has to come before the first conversation, otherwise it sounds like an excuse.

Bring a cause, not a number. "Your Thursdays are all 10 hours — what is happening on Thursdays?" is a conversation. "You worked 10.4 hours Thursday" is an accusation with no exit. Usually the answer is structural: too many tickets routed to one person, a client whose timezone forces evening calls, a release process nobody automated.

Change something you control. If you raise it and then nothing changes, you have just told them you are watching and you will not help. That is the worst of both worlds. Move the deadline, hire the second support person, kill the Friday report nobody reads.

Never make it a performance note. The moment "you work too much" shows up in a review, people learn to stop the tracker and keep working. Then you have no data at all, and the problem is now hidden. That is the whole failure mode: tracking only tells you the truth if telling the truth is safe.

The limits — and where this is the wrong tool

Be honest about what the data is not.

Keyboard and mouse activity is not attention. Somebody reading a 40-page spec for an hour looks idle. Somebody on a video call looks idle. Gaps are evidence of nothing on their own, and so is continuous input — a person pair-programming for three hours with a friend is having a better time than someone alone in a spreadsheet for two.

"Logged off" is not "resting". The person who stops at 18:00 and then thinks about the bug until midnight does not appear anywhere in your dashboard. Tracked hours catch screen time, not mental load. Some of the most burned-out people I have worked with had tidy timesheets.

A tracker will not fix a culture problem. If your team works nights because you ship on Friday and plan on Thursday, no software will spot that for you — you already know. Buying a monitoring tool to discover you are understaffed is an expensive way to avoid a decision.

If you have five people, you probably do not need software for this. You need to ask them, every two weeks, properly, and listen to the answer. Tracking helps when you have enough people and enough timezones that you genuinely cannot see the pattern anymore — around ten upwards, in my experience. Below that this is a management habit, not a tooling gap.

And if your real goal is enforcing breaks, this is not it. There are apps that interrupt you, lock the screen, run a Pomodoro timer. A time tracker reports; it does not intervene. Do not buy a reporting tool and expect a behaviour tool. If you want both, you need both.

The two-minute version

Once a week, for each person, look at: longest unbroken block, number of 15-minute-plus gaps, and span from first to last activity. Three numbers. If the block is over 5 hours or the gaps are zero, that is a conversation — about workload, not about them.

Do it weekly and you will catch things a month or two early. That is usually enough time to actually change the workload, which is the only part that helps.

If you need the raw material for those three numbers, that is what bibotracker does: it records work time, screenshots and browser activity on macOS and Windows, and shows the day as a timeline in a web admin so you can see where the gaps are — or where they are not.

I cannot get my wrists back. You can probably keep your team's.