Experiment run: 7–20 September 2026 · Published: 28 October 2026 · Last updated: 28 October 2026 · Reading time: ~5 minutes · The alarm log, the photographs and the numbers in this post are my own.
I did not need to be up at 5am for anything. No commute, no children, no job that starts early. I did it because every second book and podcast on the subject says it is the difference between a good life and an ordinary one, and I had two weeks in September where I could find out whether they were right.
I finished the fourteen days. That is the one part of the experiment that worked. Every other part of it failed, and the way it failed is more useful than the way it would have worked.
What I changed
Target bed 21:30, target wake 05:00, which is seven and a half hours. A little less than the seven and three quarters I had been getting, and I assumed that would be fine. I set three alarms and put the phone out in the hallway so I had to walk to it, which I had already been doing since the charger experiment the month before. The mornings had no structure on purpose. I wanted to find out what would happen if I gave myself two extra hours with no plan, and I told myself that finding out was the point. That decision was the whole problem.
The fourteen mornings
Week one was easy and I did not trust it. The alarm at 5am was shocking, but the shock carried me through the first hour, and I was at my desk by 5:40 with a coffee and a strong sense of virtue. That lasted about five days.
Days six to nine were the ones that broke it. I was falling asleep on the sofa by 21:40, waking at half past ten when the heating clicked off, and then lying awake until gone eleven. My average bedtime across the fortnight was 22:20, fifty minutes later than planned, and the fifty minutes was the whole failure. The alarm did not get easier as the days went on. It got worse every single morning, in a way that felt less like tiredness and more like a debt.
By the second week I had started to slide it. 05:00 on the Monday, 05:10 by the Wednesday, 05:20 by the Friday. I did not decide to do this. I did it the way you do things at 5am, which is without thinking and then feel bad about it at lunch. The log says nine mornings at five, three at ten past, two at twenty past. That looks like discipline and it is not. The slide had only just started when the fortnight ran out, and I have no confidence it would have stopped at twenty past.
What I actually got done
I gave myself about two extra hours each morning, roughly 27 hours across the fortnight, and I spent most of it badly.
On the good mornings, maybe five of them, I read, or wrote, or went for a walk in the dark. A real walk, not a performative one, because there is nobody to perform for at 5:30. On the average morning I drank coffee and looked at my phone and waited for the world to open. That is a thing I did not know I did until I had two hours with nothing in them. My honest estimate is that I used about 25 minutes of the extra time well each day, a little under six hours across the fortnight. The other twenty-one hours were waiting.
The evening is the half of the ledger I was not expecting. Before the experiment my best working hour was 20:00 to 21:00 and I used most of it. During the fortnight I was on the sofa by about 20:20 on most nights, so I kept twenty minutes and lost forty. Across fourteen days that is nine hours and twenty minutes of genuinely good evening work, and none of it came back at the weekend.
So: about six hours of useful morning time gained, about nine hours of good evening work lost, and fourteen hours of sleep gone in between. On productive time alone, the experiment cost me three hours net. If you count the sleep, and I think you should, because I have not got it back, the real price was much higher than three hours, and it is the part I would not repeat.
What went wrong
The advice is not wrong. It is answering a question I do not have.
Every version of the 5am routine I have read is written for someone whose morning is the bottleneck. Their problem is that the day starts before they have done anything for themselves, and the fix is to start it earlier. My bottleneck was the evening. I already had two usable hours after dinner and I was spending them on a phone, which is exactly what the charger experiment had already shown me and which I had not acted on. A different problem, with a different solution, one I had half-solved a month earlier and then stopped paying attention to.
What the 5am experiment actually did was tire me out so thoroughly that I stopped using the hours I already had. The problem was never the number on the alarm. It was that I had misdiagnosed which part of my day was broken.
The second thing that went wrong is smaller and more embarrassing. I had no plan for the mornings. I assumed that having time would tell me what to do with it, which is the same mistake as buying a notebook and expecting to have written something. An empty hour at 5am is not the same as an empty hour at 10am. At 10am you can call someone, go somewhere, do a thing that requires the world to be awake. At 5am the only things available are the things you already planned, and I had not planned any.
The thing I got wrong
I thought I was testing a wake-up time. I was testing a schedule, and I had only changed one end of it.
The 5am advice comes as a package: early to bed, early to rise, a structured morning, and a specific set of things you do before the world wakes up. I took the waking-up part and left the rest, which is a bit like taking the flour out of a recipe and expecting bread. The people who make this work are not doing it because 5am is magic. They are doing it because they decided in advance what the two hours are for, and they protected the bedtime that makes the wake time possible. I did neither, and I got the result you would expect from doing neither.
And the corollary, which is the actual finding: a 5am wake-up does not add hours to your day. It moves them from wherever you were spending them at your best, to a place where you are working at your worst. I did not need more time. I needed to stop spending the good hours on my phone, which is a different problem with a different solution.
Would I do it again?
No, and I want to be specific about the "no."
I would do a version of it that costs less. If I had a genuine reason to be up at 5am, a project on a deadline, a training plan, a commute, I would set the alarm and protect the bedtime like a work meeting, and I think it would work, because the reason would be doing the work the plan was not doing.
What I would not do is wake up early in order to have "more time," because more time with no plan is not more time. It is the same time, spent at 5am, with worse coffee.
Frequently asked questions
Did you actually get up at 5am every day?
Fourteen mornings, twelve of them at or before 5:10, and the last two at 5:20. I did not miss a day entirely, which I had expected to. The mornings I was latest were the ones after the worst nights, which is not a coincidence. It is the mechanism of the failure, written in the log.
What did you do about the evenings?
Not enough. I planned an earlier bedtime and did not protect it. The gap between the 21:30 target and my actual 22:20 average was the entire problem, and I did not solve it in two weeks. If I ran this again, the bedtime would be the experiment and the wake time would be the side effect.
Do you think 5am works for some people?
Yes, for people whose morning is genuinely the bottleneck, and I think those people are a minority of the people who write about it. If your evenings are already well spent and you need a quiet hour before the day starts, this is probably the tool. If your evenings are where the time is being wasted, waking up earlier just moves the waste somewhere you are too tired to notice it.
What would you try instead?
A protected hour between dinner and the sofa, with one specific thing planned for it, and no phone. That is a boring experiment and it is the one I should have run. The 5am version was attractive because it sounds like a transformation. The evening version is attractive because it might work.
About this post
Written by Maha Alkumaish. I write about the routines and systems that survive a real working week, tested on myself before they get written about. The alarm log and the figures in this post are my own.
All photographs and screenshots in this post are my own. I was not paid or sponsored by any brand mentioned, and there are no affiliate links in this post. More about me and how these experiments are run.
Related reading on this blog:
- I moved my phone charger out of the bedroom for two weeks, the experiment that had already shown me the evening was the problem, not the morning
- I did a no-spend month with three exceptions, where the fix for a rule that kept collapsing turned out to be a list, not willpower
- More Wellness & Routines experiments
0 Comments