The biggest thing I learned when analyzing my time, which I wish I had recognized earlier, was that I can have two types of productive days:
I can have days where I do a million small things and hop quickly from task to task.
Or, I can have days where I work on big issues and should not be interrupted by small issues.
But, if I think I'm going to have a day where I work on a big thing and it turns out I get interrupted by a million little things, I end up doing nothing well and end up very unproductive.
This is why I and most of my team work from home on Fridays (we never book meetings either). It's tempting to dismiss working from home as a way of skipping on work, but there's no other time I can get 4+ contiguous hours in front of my computer with no interruptions or meetings. It's probably my most productive dev time all week, largely because I can plan for it. I frequently and vocally encourage work from home on Fridays to become a standard (but optional) "perk" in tech.
I used to take the shuttle to Twitter for an hour each way. I think I wrote most of the code I wrote there on the bus because of the reliable time in front of the computer.
Yep. Mirrors my experience with shuttles. I, too, have banged out some real work on a shuttle with headphones on. I unexpectedly regretted moving closer to the office because I lost this time, despite that making not a lick of sense.
Offices are for managers to feel that people are productive, not for people to be productive. Look around most offices and there is absolutely no space laid out for concentration time, except social concentration time.
> But, if I think I'm going to have a day where I work on a big thing and it turns out I get interrupted by a million little things, I end up doing nothing well and end up very unproductive
These can be some of the most frustrating and dispiriting days
I found other method to be effective for me. I work from home till noon - this are the best "brain hours" for me, so it's perfect to dive deep into hard dev tasks. I arrive at the office at ~12 and do all the biz stuff. Works good so far.
It might just perceptual: it might be the case that days very into on of the two types are easier to recognize, tag and remember, whereas mixed days are not so vividly remembered.
I surely stop after a day of lots of little tasks and say "phew! What a day!" or after a big task is done "yeah! look at this cool feature you built today!" And even play with it a little. Normal days are not so reflected upon...
Very good question, I thought many times about this and could think that probably the beginning or how you start the day determines the rest of the day.
The notion that a 130h work week is admirable, desirable, sustainable or useful is ridiculous and should be criticised every time it's raised.
Unless you're working for yourself or working in a job where your contract compensates you by the hour then investing such huge swathes of time I think is destructive.
There is a reason workers united, fought and were martyred for the 8 hour work week and the creeping clawback by industry is a problem.
That aside, very interesting to see such a consistent time keeping record.
While I agree with your statement for the most part, I also understand that there is some context to it. That specific context is Marissa Mayer and other early Google employees. I find it hard to convince myself that those people and especially Mayer, were not or are not compensated in some way that takes those huge swathes of time into consideration.
She was paid $36MM in 2015 and raised the market cap of Yahoo! by 151% in four years. Some of that compensation is based on success at Google, some of it may be based on her success at Yahoo!, but some of it is absolutely based on her brand. Her brand, in my own mind, is highly associated with her dedication to working long hours.
I work long hours but I'm conscious of a few things: I don't celebrate it. I don't plan to proselytize it like Mayer does. Part of those long hours are dedicated to operationalizing time-sinks so that I don't have to do it forever. I don't expect others around me, especially non-founders, to do the same and like you I believe that for them to do so would likely be a net-negative.
Working long hours on the basis you might be compensated if the company goes on to be worth billions is akin to buying a lottery ticket. For every Google there are thousands of examples of businesses that achieved far smaller success or failed.
If you want to work long hours, do it because you enjoy the work. That way there's a reward now.
I was under the impression the majority of the change in stock price for Yahoo(if not all) was related to them owning a large share of Ali Baba stock which has skyrocketed in value. And this purchase was made before Marissa Mayer took over.
This could be the case. I wasn't able to quickly find market cap information for Ali Baba prior to it's US listing. Every since the Ali Baba IPO, Yahoo!'s stock has closely tracked BABA.
Do you think that market cap of Yahoo was increased by 151% in four years as a direct result of the number of hours worked? Could it have been done if she worked 50% less hours? A difficult question to answer with evidence.
> Do you think that market cap of Yahoo was increased by 151% in four years as a direct result of the number of hours worked?
No.
I believe it's probably true that she is compensated, in part, because of her brand which is highly associated with her dedication to working long hours.
I hedged quite a lot in my OP to try and convey that this was the only part that I was confident of.
I also for a second don't believe it is productive time.
I'v worked similar hours, and believed I was productive, until I had a team and saw what happened to their productivity when they worked long hours.
Not just did it damage their productivity, but it killed overall team productivity, because it caused a rise in conflicts, rise in "downtime" where people were physically present and in theory working but didn't get anywhere.
This might "work" for some notions of work for simple repetitive tasks where you can have someone overseeing the work and "whipping" people to keep tempo up.
I have never seen it work with more than very moderate, very temporary, increases in working time, across a large number of teams over the last 20 years.
In fact, I'd argue that a 40 hour week is pushing it for developers. I have my most productive weeks when I work less, and I know many who are the same.
Thanks for sharing this. Is the metime code publicly available? I'm the CTO of another small tech company (https://sourcegraph.com) and I'm also a bit obsessive about time tracking.
I wrote a small open-source CLI that gives you a CPU-profile-like view of time spent on your computer: https://github.com/sourcegraph/thyme. Thought I'd share since OP and others here might find it useful, and I'd love to hear any feedback.
Hi author. Small nit: consider changing your title to "their" time rather than "he" since you talk about "a technical co-founder". There's no need to use gendered language in the general case, and gendered language needlessly encourages gendered based identification. Imagine if you included your race and ethnicity and sexuality in the title - that'd feel silly, right? It would limit the audience who found the piece relevant, and it could be read as though those aspects were somehow relevant to being a technical co-founder. No doubt gender and race and ethnicity do effect one's experience and interactions, and that's a great piece to write, but that's not what you're trying to talk about in this piece so why put it in the title?
Thanks for the feedback on the title. I've changed it to "their", which still works fine for me.
For the avoidance of doubt, I do say in the article that:
> The CTO role is a very diverse one, and I don’t expect that the results here will be true in general. This is just my story. I hope it is an interesting one, and it is uniquely backed by data!
So, the subject is me, and I am a male. I did not intend to exclude anyone by using "him".
(I can't change the title on HN myself, but if someone who can wants to do so, please feel free.)
> Imagine if you included your race and ethnicity and sexuality in the title - that'd feel silly, right?
Historically it has been correct English to use the masculine pronoun where it could refer to either sex. I understand the reasons to prefer a different construction ( I also find "their" or "one's" to be preferable) but comparing standard English to explicitly calling out ethnicity or sexual preference is virulent.
Off-topic: I really like Overleaf. It removes the barrier of installing all of the compiler, editor and dependencies and allows anyone to jump right in and and start typing manuscripts. That's why I always recommend Overleaf to anybody looking to learn Latex initially.
It's interesting to note that the meetings time decreased as the management time increased, because often the two are strongly correlated.
In my experience meetings and management time grow proportionally, since meetings are a good way to talk with the team, set expectations and review the results. Or, in other words, some meetings == management.
> My app had some simple charting built in but no real analysis. It’s only now, six months later, that I’ve had a chance to really get into the dataset
I'm interested to find out if the data influenced the OP from week to week (month-to-month) during the collection phase? Did it influence his keeping up his development by shifting it to weekends? Or was there other catalysts to that? What were they? Were there planned development milestones? were they sized?
(Author here.) The shift to developing on the weekends wasn't really evident in the simple charts that I had --- I only found out about that after doing the detailed analysis, and it required doing some statistics to really pick out the trend from the noise.
I have now open sourced the app, if you'd like to dig deeper into what I was looking at during the (latter part of the) experiment: https://github.com/jdleesmiller/metime
We did have some large projects to complete on tight timescales that spilled into the weekend. We didn't really have a good system for sizing / pointing tasks (and we still don't) --- probably an area where we could improve.
Awesome project! I'm a big fan of using econometric methods in Quantified Self -type analysis.
A couple of thoughts on your regressions:
Keeping holidays in the data seems counterintuitive. I'm fairly sure just from eyeing the graphs that the increase in your work week might reach statistical significance if you were to use a regression method more resistant to outliers, like Least Trimmed Squares.
Another idea - in microeconometrics it's standard to use some type of Poisson regression four count data, which this could potentially qualify as.
And furthermore, I would have loved to see some sort of distributed lag model of the timeseries (or at least a scatterplot of all activities against each other) to see which ones tend to co-occur.
(Author here.) I'm very pleased to finally receive some feedback on the technical side!
I will have to check out Least Trimmed Squares (and maybe other robust regression techniques). For this dataset, I did try to exclude holidays, but it was difficult to define this exactly, so a more robust technique might indeed work better.
And yes, it would be interesting to see what tags co-occur. Good ideas for a future post. Thanks!
I work as a freelancer (Europe) and I think I work reasonably hard. But is really difficult for me to honestly log more than 10-15 dev-billable hours during the week. The rest of the time is spent in chat and learning / playing with a new things.
Do your clients understand that playing / trying out new things leads directly to you building the finished product?
Or is your playing truly separate from any of your client work?
My max was 48 hours/week ( but this is an exception )
For living I build e-commerce solutions and backend stuff. And in free time I play more with a frontend. Recently with react/redux and in the long run with every possible new technology on the horizon. I guess I waste a lot of time doing this as most of these things isn't immediately related to client work. ( but over time It helps me to sell to them a few ideas for improvements )
Also, I track my time and habits in detail ( using rescuetime ). Probably without such careful loging I would overestimate how much time I am spending on actual coding.
I live in Warsaw and live here is a lot cheaper than in Western Part of Europe. My clients are Canada / UK based and I charge between $50-$70/h so even 10-15 hours a week is enough to make living over here.
Interesting article, pretty remarkable to be that committed to the time tracking -- I can barely stick with using the pomodoro technique for longer than a few days in a row...
I hope OP continues to grow the dataset. Will be interested in a follow up later on!
I recommend not only track your time, but also write down what you do, and have project-names. Using this I can give very accurate estimates by going back and see how long time things took. And it makes it easier to plan things.
Also if you are in an early startup it's good to have this so your co-founders know how much you work and on what.
And if you are employed you can use it to tell your boss exactly what you did (and why it took so long).
There's a saying though, that you'll spend 50% of your time documenting what you do.
I can have days where I do a million small things and hop quickly from task to task.
Or, I can have days where I work on big issues and should not be interrupted by small issues.
But, if I think I'm going to have a day where I work on a big thing and it turns out I get interrupted by a million little things, I end up doing nothing well and end up very unproductive.