Comment by DanielBMarkham | Hacker News Reader

Comment by DanielBMarkham | Hacker News Reader

I've worked in every possible configuration, and I've watched teams work in every possible configuration.

I don't like open-plan, but it tentatively appears to be most effective for organizations as a whole. Some initial conclusions:

- We'd be wise to distinguish what we'd all like to have as an office from what actually functions best in a larger system. It is erroneous to think that simply because you like something means it's always the best policy for everybody. Sometimes local non-optimizations need to be made to optimize the system as a whole.

- Open-plan environments are nowhere near being a determining factor in team performance, but they can be an important add-on. So sure, if your company is too cheap for offices, treat teams like slaves, and just wants to cram as many people possible into the smallest space? Then that's going to suck. But that's going to suck anyway. I'd just be really careful about confusing "stuff I think is cool" from "stuff that works the best for everybody concerned" I'd also be careful about over-generalizing. A big hunk of this is team, project, company, technology, and building specific.

- As long as you're adapting, you could end up with any kind of office space. But just as an informed guess, co-located open plan is where I'd start, then adapt from there.

- The worse thing you can do is start with a pre-conceived notion of what works and didn't and then try to force reality to conform to it.

"I don't like open-plan, but it tentatively appears to be most effective for organizations as a whole."

The original article is quoting peer-reviewed research that tends to support the opposite of your opinion. What do you think is wrong with the methodology employed in the OA?

For my part, I wouldn't necessarily disagree with the article (and I would need to read through the whole paper to know for sure), but I agree that most teams probably don't function well with open concept offices, but imo that's likely a function of how (in my experience), lower-functioning teams are probably made worse by unfocused background conversation.

Again, my sampling is based on the good fortune of working on really good, highly talented and focused teams. It's not perfect and certainly after months and years on the same team, you can lose perspective and think some things are worse (or better) than they really are. But I've been on really disfunctional teams as well, and some teams that I'd say were "in between". For those latter teams, I wouldn't recommend open-plan. To me, open plan is a high-risk, high reward proposition that takes commitment from the top down to make work over the long haul. But in my experience, the reward is high, and I really do prefer it.

I can help with that:

"Workers in open-plan offices get sick more often" - Only supported by one study done in Denmark. Not enough research.

"Open-plan offices are less productive" - Supported by a study which only identified some of the negative effects of a open-office plan (such as the effect of irrelevant speech). That study didn't attempt to measure the benefits of an open office, or attempt to exhaustively list all of the effects of an open office, so it doesn't speak to the overall effect on efficiency.

All the other claims were either focused on employee opinions or were about things other than an open office plan.

You can guess what my opinion here is. At my previous company, the floor plan had both offices and a wide open area, so we could choose where to sit. I picked the open area and loved it. Being able to eavesdrop on the status of other teams, or having really easy group communication (as easy as yelling across the room) was incredibly valuable. I think the open office is especially nice in our industry since we really tend to suffer from under-communication and a lack of cross-team networking.

Anyway some folks at our company chose small offices and I think they enjoyed it. The ideal floor plan would have a mix of both I think. But I don't think Facebook's new one-giant-open-area building is nearly as awful as people make it out to be.

On the other hand, I'll trust even one peer reviewed well done study over someone's passionate anecdote any day. Even the best of us can fool ourselves into believing incorrect things. Experience doesn't help in this either. Unless, of course, your experience involves doing proper testing of ideas. But of course, then you have data.

I worked in a similar, one-giant-open-area office for a year and didn't enjoy it.

My team was a cluster of 8 fixed desks in a giant sea of engineering, with people always walking through our area on their way somewhere else. The noise level was so high they installed custom white-noise generators that have an unnerving effect when you spend all day with them.

Contrasting this to another company where there were separate team rooms of 5-10 people for each group in the company and it was much more pleasant while still having a lot of the same benefits.

Couldn't agree more. What people complain about is not necessarily an indication of what's not working.

My last two gigs have been open-plan, and I prefer it much more than closed-cube spaces. If your team is working as it should, and is in close proximity to project/product management, with a good process and a healthy level of trust between the stakeholders and between the various engineers, then in my experience open-plan is far better, because the quality of communication is high, the general focus of the team is high, and there's less background noise "distraction", because the cross-talk is usually relevant and on-task, and has a higher likelihood of being pertinent to the job at hand. Open plan offices work well when the team is already working well.

However, if there's broken processes, unfocussed coworkers, and a general level of distrust between team members or across teams, then the background chatter becomes exponentially more distracting and frustrating. Unfortunately most teams tend towards the latter and not the former.

In other words -- in my limited, anecdotal experience (last 5 years, two large open-plan teams; 10 years before that in closed or hybrid), high-functioning teams tend to thrive on the cross-talk and constant communication that open plan creates organically. Lower-functioning teams on the other hand have done better when the natural barriers of walls and closed doors limit unnecessary distractions and create healthier boundaries.

So it's not surprising to me that the study shows more instances of teams unhappy with open plans; it's certainly not for everybody. But the best teams I've been on have absolutely thrived in that environment.

> high-functioning teams tend to thrive on the cross-talk and constant communication that open plan creates organically.

I wonder how programmers in such environments can get in the "zone" and stay in it for extended periods of time, if they are being "constantly communicated with".

I'm in a team that benefits from an open space and frequent chair swivel initiated conversations. Thankfully my chair has wheels and my laptop a headphone jack because it's not terribly difficult to go sit in a corner.

This is a key question. Also everyone considers two alternatives, open and cube, forgetting that individual offices is also an option

I'm ADD to the core, and I find it actually easier to zone out and code with a low-level background noise, "coffee-shop" style. Large spikes in noise are a problem, and also non-relevant conversations that pull everyone's attention off-task, but on the open teams I've been on, in general this has been the exception not the rule.

My experience at large tables has been that good coders and PMs will actually IM a developer when they need to ask a question -- I do this all the time even when I'm sitting 2 feet from the guy I need to talk to: "let me know when you have a sec to talk about X", etc.

Again, that's evidence of a team that's generally on-task and communicating well -- a cultural quality that is itself harder to find than not.

I prefer instant messaging to start a conversation... it's asynchronous. You can ignore it for a little while if you need just a couple minutes more of flow state

This has worked well for our teams. As a social norm of software development I always IM/email someone with their headphones on and don't interrupt them.

It's also a nice courtesy to IM someone before dropping by their desk so they aren't surprised and you can check if it's a bad time.

Don't try to think large complicated thoughts in your head, think them written down on paper. Typing doesn't work as well (as far as I can see), I would guess that a large enough tablet with a stylus would work better.

This will increase your capacity for involved/interesting ideas, make recovering from interruptions faster, and slow you down slightly (but not enough to completely offset the benefits).

In my experience, exactly the opposite can happen and open-plan offices can discourage a lot of that cross-talk and casual discussion. The last two I've worked in had an almost library-like atmosphere.

In a quiet environment, people tend to be discouraged from being the only ones talking and being listened in on, and instead scurry off into the meeting rooms or manager offices. Others put on the headphones and go totally 'heads-down'. Communications end up being more closed and 'formal'.

Step 1: Get your team working "as it should."

I don't see much difference between open plan and high-walled cubes in a large room. Noise and distractions are about the same. I think if you're going to talk about private space vs. open you have to be talking about real offices with walls from floor to ceiling, doors that close, and which are (reasonably) soundproof.

Sorry, should have changed that to "offices", not cubes. Haven't spent much time at all in cubes, and the couple of months I did it was the WORST of all worlds.

"Noise and distractions are about the same"

I found interrupt rate increased for open plan by, oh, probably a factor of ten. Look, VLM isn't talking, therefore he must not be doing anything, lets ask him about last nights TV episode. If the cube wall is high enough that you can't see if the worker drone is in there much less on the phone (aka working) or actually working as in thinking, then you don't just yell at the guy across the room.

I notice a massive impedance bump between the "knowledge workers" and the "talkers". Engineering and Finance and operations got along nearly perfectly because both know that not talking does not mean not working aka ready for interrupt please bug me. On the other hand management / customer service / sales, if they're not talking they're not working, so interrupt rate was horrible from those types, hmm I have a problem and those three are talking about football so they're busy, and that guy's talking to his wife on the phone so he's busy, I know I'll bug that guy looking at a component datasheet, obviously he's the only one here not doing anything. Marketing types were kind of in between, some understand that thinking is work, some don't.

I see office environment as a proxy for willingness to expend effort and time in making high performance teams and letting them get on with it.

If a management tier is unwilling to spend money arranging desks to avoid constant interruptions that common sense tells us will be detrimental, then it is a management tier that is unlikely to have been a blip and then make all other decisions well.

Google Ed Catmull, the quiet co-founder of Pixar, interviewed at economist conference 2010 - it is far and away the best example of simple focus on getting teams to work well - no "do it my way" but plenty of "if the team is functioning well there are enough good people in it to over one anything"

I'm not sure we're disagreeing.

I'm simply saying that performance -- as measured by the team itself -- may be maximized by the team deciding to do things that it finds uncomfortable. You also don't always optimize overall performance by optimizing individual performance.

There's also the issue of communication versus productivity. Communication is a far greater obstacle to many technology team's success than individual productivity. Bunch o' guys working by themselves in a nice, quiet cushy office doesn't give you the cross-talk, annoyance, problem-solving and innovative banter you get by putting them all together. Sometimes doing things you find annoying or troublesome is actually the best way to reach your goals. Surely this isn't a new observation.

Agree on the communication (I say honesty - you can communicate all day but if you are not feeling therapy-level angst about saying the things you are saying out loud, then no one is getting to the root of the problems and they will never get dealt with)

Not too sure about the optimising for individual vs team performance. Teams seem to do best when it's obvious everyone has to play their very best game. The times when it's necessary to sacrifice for the good of the team are a hell of a lot less frequent than the times you just need to know task A will get done because Fred is on the case and Fred has a history of delivering. Stops everyone trying to worry about everything.

Edit: the sentence I'm not sure we are not disagreeing either does not make enough sense to keep.

Reply on news.ycombinator.com