A few weeks ago, my group participated in a mandatory work from home day. This is an ongoing thing that all groups in the central IT organization are doing in turn. (The reason why is not entirely clear to me. Cynically, I suspect it’s because they want to make people work on snow days or they want to oversubscribe our office space. That’s entirely speculation on my part.) I’m not convinced that test is very useful, since it does little to simulate real-world conditions. Knowing ahead of time that I’d be working from home, I took steps to ensure I was prepared.

Since I normally leave my workstation on anyway, I mainly had to take home my headphones. The rest of my work could be sanely accomplished over SSH, through replicated files (via SpiderOak), or with a web page. My group already uses a Jabber server for IM communication, and we coordinated a Google+ hangout in order to conduct face-to-face meetings. (As an aside, I’ve developed a new management maxim: you should never have more direct reports than can fit in a Google+ hangout.) We worked fairly well that day, at least from my perspective. Since it was a day when no meetings were on my calendar, it was a fairly reclusive day anyway. I did end up having one meeting with someone outside the group, which we conducted via Google+. It worked well enough. A better test would have been conducted on a more meetingful day, with the entire IT organization home at once. Really, a multiple-day test would be better, since I suspect the temptation would be to put off complicated interactions until the following day.

This morning, faced with the possibility of a very unpleasant commute home (near-blizzard conditions existed most of the day), I opted to work from home. This time, I didn’t have my headphones, which wasn’t much of a problem because there was nobody to call. So many people take the days between Christmas and New Year’s Day off that even if I were in the office, it would basically be a ghost town. It was nice to see that an unplanned WFH day worked about the same as the planned one, though it remains an un-representative case.

I’d like to see this test conducted in a more meaningful manner. As I alluded to above, a full week with all of the IT staff working from home would provide a more meaningful result. My productivity was slightly dependent on my workstation at the office remaining online, though I could recover from that if I needed to. It would just involve finding by accident all of the packages I’ve installed at work that I don’t have installed at home. I’m interested to see if there are any follow-up efforts.