Productivity – A Data Based Introspection

The other day, I happened to remember that I track the seconds spent typing into my code editor, and have been since the beginning of 2016. It’s nothing spectacularly special, but I wonder what I would see if I pulled the data?

I put the data in excel and used a 14 day rolling average:


The average number of seconds isn’t important, but the trends are. You can easily spot that I have 2-3 week spikes every couple of months. Those are likely projects getting started and since we’re counting seconds typing, I guess I have a lot to type. What’s more important is that, lately, I’m typing 40-60% of my regular amount. So, I’m not as productive as I used to be! What happened?

Well, I’m still learning a massive codebase and tackling problems that require actual research, I’d expect to be typing into a code editor less. However, are there any other things buried in the data?

I pulled the server logs and got the ip addresses and then ran a geoip tool on them. I was able to deduce where I was and when. Based on that, I learned a few more things about myself.

Looking at the graph, you might spot a slow downward trend over the last few months, this correlates to the amount of time spent at my coworking spot, where I have two monitors. The last several months, I’ve found myself working at home more and more. In fact, every spike in the past correlates to a place where I have more than one monitor. It has nothing to do with projects at all. Naturally, every drop correlates to my travelling and/or vacation.

So, Monday morning, I cancelled my coworking space and brought my other monitor home and set up a proper home office. It has been amazing and I seriously wonder why I never did it before.

When you throw in an extra computer monitor, it comes out to (on average) at least one standard deviation above the mean in extra typing time (in other words, a couple extra hours per day!). I’m starting to wonder how much this scales! Maybe I’ll type more if I get yet another monitor?

Maybe something like this?

Surprisingly, there’s days where I barely type at all. Going through various resources, it looks like I was whiteboarding or debugging. There’s plenty of 0’s to go around as well, either due to holidays, forgetting to install my tracker on my new computer, playing too much Playstation, or parenting.

The impact of a secondary monitor is probably not all that surprising to some people, but it was to me. So, when using my multiple monitors over the last couple days, I paid attention to how I was using them. I noticed I kept my code editor and Chrome in my primary monitor. I switched between the two windows just as I did with my single monitor. In my secondary monitor I kept Slack, IRC, and my terminal open, but usually slack was in the foreground unless I was actively tailing logs.

By having Slack in the foreground, I could quickly scan a notification/ping and triage my attention and go back to what I was doing by just shifting my gaze. When using one monitor, I wouldn’t triage. I would just shift my entire attention and deal with it, whatever it was, right then.

Actually, I think this is the ultimate blessing and curse of working remotely. In an office, someone can, and will, physically interrupt you if they think it’s important enough. Working remotely is a blessing because it’s easy to go “heads down” but a curse because sometimes a person doesn’t respond quickly enough (myself included 😱).

I don’t quite understand how having multiple monitors can account for two hours of productivity. That seems like quite a stretch, but it must be the largest factor based on correlation. It could also be due to how I count “typing.” I count it pretty simply: pressing a button in the editor starts the clock. Once 15 minutes passes without any typing, it calculates the last button press to the first button press and sends that time to the server. If another button press happens in that 15 minutes, it counts the intervening time as “typing”.

In other words, with multiple monitors, I can have a conversation without fullscreen context switching, while on one monitor if I’m expecting a reply (“several people are typing?”) I won’t switch back to the editor. So, maybe I can see how having multiple monitors can make such a difference. I guess only time will tell.

I’m glad this foray into my personal tracking data has led me to bring home my monitor and set up a legit home office. It’s allowed me to move faster and break more things the last couple of days. I like breaking things. I’m interested if you’ve noticed the same things, I expect it to be different for different people. I’m curious just how different it can be.

PS: I’d like to thank Past-Rob for setting this up and working in strange places so that I could draw the correlation between monitors (instead of workplace)!

Want an inside scoop?

Check out PHP Shenanigans for only €5/mo or €30/yr