I've Kept a Work-Time Spreadsheet for 2 Years. Here's What It Showed
For two years I've kept a spreadsheet where I log every workday. Date, hours, what I did, the plan for tomorrow. At first it was a tool for invoicing clients. Over time it turned into the most honest data source about my work. And the main takeaway was unexpected: if I work more than six hours a day, my happiness level drops.
How the habit started
When I started taking on work as a developer, I needed to tell clients how many hours I'd put in. At first I eyeballed it, and I quickly ended up on both sides of the problem. On some projects I worked for free past the estimate; on others I felt guilty for charging a lot for something I did fast.
The solution was simple: a spreadsheet. I open it in the morning, note the start of work. Close it, write the hours. At the end of the day, one row. Date, hours, what I did, what I plan to do tomorrow.
Six months in, I realized I wasn't using it just for invoices. It started showing me a rhythm: when I'm productive, when I burn out, which tasks really eat hours and which only seem long.
What's in the spreadsheet
Four columns, nothing extra.
- Date. One day, one row.
- Hours. Decimal, not "10 to 6 with breaks," just "5.5".
- What I did. A couple of lines of free text, projects and tasks.
- Plan for tomorrow. The main action of the next day.
No categories like "focus time," "deep work," "meetings," "rest." I tried it, and it turns into a process of its own and the spreadsheet itself stops living. The fewer fields, the longer you keep filling it in.
I count hours strictly. The time I'm in code, in design, or on a project call. A coffee break and scrolling my phone don't go in the spreadsheet. A walk while thinking through a task does, if I actually came back to that task afterward.
The numbers over two years
Over two years the spreadsheet has about 600 workdays. The average is 5.2 hours a day. My longest unbroken sprint was 9 days at 8+ hours. That was release week for one of the client apps. Final feature polish, App Store review, bug fixes on launch day. After that sprint I just lay around for three days.
The shortest workdays are 1–2 hours. Usually a Monday or Friday with meetings or admin work. Sometimes the day after a big sprint.
Days with zero hours don't go in the spreadsheet. Those aren't "skipped days" — they're normal days off.
Where the tipping point is
After six months of tracking I noticed a strange pattern. On days when I worked 8+ hours, I was tired, irritable, and usually unhappy with the result. On 4–5 hour days, the opposite: I'd wrap up feeling "I got it done, and did it well." This isn't a correlation with task type. Eight-hour sprints happened on both simple and complex tasks.
I started noting my mood at the end of the day in the spreadsheet as one line: empty, so-so, bad. A couple more months and the picture was obvious. Past the six-hour mark, my subjective state starts to fall. Not a landslide, not all at once, but steadily.
At seven hours I can still work productively. At eight I'm already getting stuck on simple tasks. At nine I write code that I'll be rewriting myself the next day.
Why four hours is the ceiling
Four hours a day means four hours of full concentration. No switching, no chat, no checking email. Sounds like little. In practice almost nobody works that much.
When I say "I work eight hours," what I actually mean is:
- 1 hour of real concentration in the morning
- 30 minutes of reading chat, replying
- 1 hour of focused work
- 1.5 hours of meetings and switching
- 2 hours of work in "hands are moving, head isn't really in it" mode
- 2 hours of "finish it in the evening because I didn't get it done"
Eight in total. Real concentration inside is two or three hours. The rest is the illusion of work.
If you strip out the switching and context switches and leave a clean four hours, I get as much done as in eight hours of mixed mode. Sometimes more. And the main thing is I come out of those four hours feeling "ready to get on with my day," not "I'll go to bed at ten because I can't anymore."
What I stopped doing after this discovery
I stopped working in the evening. If I didn't get to it during the day, I moved it to tomorrow. No "I'll finish it after dinner." I think poorly in the evening, and all the code I write at night I on average rewrite in the morning.
I stopped filling time with "not that important" stuff. Before, any gap in my schedule got filled with small tasks. I'll answer emails, google a library, look at someone else's code. Now I know I have a finite resource of concentration, and I don't spend it on things I can delegate or not do at all.
I stopped "working Saturday because I was lazy on Wednesday." It always ends the same way. Sunday vanishes into recovery, and on Monday I'm a wreck.
I stopped being proud of "working 10 hours." It's a symptom, not an achievement. If you're working 10 hours, you either misjudged the task, don't know how to break it down, or you're trying to compensate for something (decomposition and estimation is a separate skill; I wrote about it separately).
What I started doing instead
I give my mornings to my most important task. The first two or three hours of the day are the most valuable resource I have. I don't answer chats, I don't open email, I don't check Telegram. I just work on one task.
One day, one big task. If a day has a feature plus a bug plus documentation plus a call, I know in advance I won't get through all of it. So I pick one main action, and the day is built around it.
Stop after six hours. Even if it feels like I'm "almost done." Almost done will take another two hours, and I'll do it slowly and badly.
Every two weeks I look at the spreadsheet. Not every day, not for micromanagement. Once every two weeks, a big-picture look. Where things were bad, where they were good, what the common thread is. It gives you a perspective you can't see from inside a single day.
When overwork is justified
Sometimes you need to work eight to ten hours. The final sprint before a release. An urgent bug in production. A client demo in two days. I'm no puritan — crunch happens, and there's nothing wrong with that.
The problem starts when overwork becomes the background. When the "urgent sprint" is in its third week straight. When eight hours a day is the norm, not the peak. That's when the spreadsheet shows you something simple. You don't have a project overload problem — you have a systemic one.
And it's also a tool for being honest with clients
A bonus I hadn't counted on. When a client says, "but we agreed on a week," I open the spreadsheet and show them. This feature took 14 hours, this API took 6 hours, the revisions took 4 hours. There's nothing to argue about — the numbers were recorded as we went.
It works both ways. If I estimated a task at 20 hours and spent 12, I'll tell the client that part of the budget was saved. That has never been grounds for a retroactive discount, but it builds trust that pays off on later projects.
What to take away from this
In short:
- Set up a spreadsheet. Four columns, nothing more.
- Count only real hours of focus, not "time spent at work."
- In two or three months, patterns will show up. At first you won't see them. Then you'll see them all at once.
- Don't buy the myth of the eight-hour workday for a solo developer. Four hours of full focus gets you further.
- Bragging about long hours was the norm in the 20th century. In development, what matters isn't how long you sit there, but what you've managed to think through.
The spreadsheet template is simple — I can send it to you on Telegram. And if you have your own time-tracking experience, tell me about it; I'm interested in other approaches.
What's next
Want to test your idea faster than in a year and a half? Estimate your project budget or get in touch and we'll figure out how to fit your hypothesis into realistic hours.


