If one person is wrong, they’re wrong. If a lot of people, some of whom got extremely rich off of their wrong ideas, are wrong, there’s a good possibility I’m the wrong one. At a minimum, it’s useful for me to understand where I’m differing from others. Open offices are one such puzzle. To me, they are obviously one step short of Azkaban. And yet everyone, including some exceptionally profitable companies, uses them. Why?
[I’m going to restrict myself to tech companies because that’s what I know]
One possibility is some people genuinely prefer them. I keep talking myself up to that, only to read another article about how everyone is miserable and unproductive in them. I talk myself up again, and find a peer reviewed study detailing their terribleness. I thought maybe they were for extroverts, but then I heard extroverts complain they couldn’t get any work done in them either (although they were having a lot more fun not working in them than I was). My friends’ defenses of them/explanations of how they make it work sound more like Stockholm Syndrome, or at best the way I sound when I find a shortcut to finish a useless but mandatory 30 minute training in 5 minutes. I noticeably improved my situation relative to the 30 minute scenario, but that doesn’t mean those 5 minutes were valuable. But let’s assume my friends are a non-random subset and there are people who thrive in (some) open offices. That’s great, if you hire those specific people. One of my major frustrations with my current employer, Stark Industries, is that their interview process (closed room, no distractions, puzzles to solve on your own) is designed to filter in exactly people like me, and the work environment (completely open, constant distractions, work that sometimes feels more like being a PM* than a programmer) couldn’t be better designed to make us unproductive.
One possible justification for open offices is cost. I certainly think that’s a larger factor than many companies admit, but if that were the only concern they’d convert to entirely work from home. Moreover, engineers are really, really expensive, and making us less productive is costly. The extra space necessary for doors or cubicles could easily pay for itself. A slightly different explanation is that even if companies were willing to buy doored offices, acquiring office space is lumpier than hiring. Having more than you need is expensive and it takes time to ramp up after a hiring spree. That could explain temporary open offices, or roommates, but not stable ones.
Let’s go back one paragraph. The open office isn’t the only thing I dislike about Stark Industries. I’m also continually baffled by the fact that my technology company has a workflow designed around synchronous communication, in person if at all possible. No one has time to answer email or IM thoughtfully because they’re running from one meeting to another, so if you want a response from someone you schedule a meeting. The correct response to someone ignoring your e-mail is to ambush them in the hallway or, if they’re at a different site, schedule a videoconference. It took me a very long time to get this, but making a meeting to do something that could have been handled over email is not a failure mode at Stark Industries. This is how they expect it to work. This must be how they want it to work, because instant messaging is a strictly easier technical problem than helicarriers, a project we also do. Information is exchanged at meetings, which means everyone has to process it at the same time and either everyone moves at the speed of the slowest person* or you leave them behind.
What if the open office and the synchronicity are not a coincidence? If you believe synchronicity is helpful (which Tony Stark clearly does, and which I agree with in some instances), then you’ll want to encourage it. But as noted above, this is not the natural mode for a wide swath of programmers. You can hire for it at first, but eventually that cuts you off from too much talent. Any one individual can be forced to switch modes by being embedded in a group full of the other, but there aren’t enough synchronizers to absorb all the asynchronizers.
But… as much some people like retreating to do their own thing, they also like it when other people respond to them immediately. They may be held back by empathy, but they’d still like the answer right away. In an open office, the barriers to demanding an answer are reduced. For one, you don’t have to leave your chair. For two, offices and even cubicles have a sense of personal bubble. You wait to be invited in, and it’s expected you’ll have to wait until they reach a breaking point. After extensive experimentation I can tell you there is no way to generate that bubble at Stark Industries, and I assume open offices in general. I once had a co-worker poke his head into the conference room I was hiding in for the sole purpose of asking if I was hiding so I could concentrate.** Open offices also lower the cost to any one interruption. They do it by interrupting you so constantly you never get into a groove that could be interrupted, but they do technically lower it.*** So even the highly empathetic will feel less reluctance to interrupt co-workers because they are correctly calculating a lower cost to it. In high doses, perhaps mixed with morale events and a culture that emphasizes meetings over email, this could lead to teams made entirely of asynchronous workers forcing synchronicity on themselves.
What is it about synchronicity that makes every major tech company started in the last 20 years be willing to pay so much for it? Based on every survey ever and the coding wars study, it’s not improved performance at the object-level tasks of the job. But work isn’t school, there’s more to it than fulfilling the terms of the assignment. Maybe open offices lead to less redundancy or wasted work. Maybe they make charisma and personal connections less important. Maybe they’re the best way to force programmers to share information in the face of their steadfast refusal to write anything down. That not only makes people more potentially more productive, it makes them more replaceable.
None of this makes me love open offices. For one I’m pretty sure I’m better at synchronizing via technology than speech. By a lot. I love Slack because it gives me everything everyone said I would get from open offices, without any of the costs. It gives me a sense of control and in-touch-ness that makes me want to read it. Meanwhile I approach co-workers in person less now than I did when we all had doors, because I’m hyperconscious of impinging on the other people in the room. But I will say I started doing better at my job when I acknowledged that I was expected to do it synchronously and rolled with it. Matching the office work style turned out to be more important to productivity than matching my own. It exhausts me, but at least it’s the exhaustion of having worked really hard. When I tried to work asynchronously I came home exhausted from doing nothing, which was a much worse feeling.
*Project/program manager. Job description depends heavily on the team but one of their jobs is to coordinate people with subject matter knowledge.
*Slowest doesn’t mean dumbest. They may very well take longer because they’re thinking more deeply.
**The answer was yes.
***The economic term for this is bee sting theory. You’ll work really hard to avoid your first bee sting, and you’ll pay a lot to get rid of it. But when you already have 10, the work to avoid an 11th just doesn’t seem worth it.