Ending the open office epidemic
wildbit.com
wildbit.com
Some qualifiers:
- The team has a good culture of respecting each other, and we're busy enough not to want to banter too much outside of certain times (e.g., lunch, when we see it already going on, etc.).
- I've got good headphones hooked to a 24-bit-DAC'd USB amp. I don't hear anyone unless I'm trying to.
- My screen's big enough where visual noise isn't an issue.
On the upside:
- Asking someone a quick question is really, really quick.
Downsides of my prior each-their-own-office experiences:
- I sometimes felt a bit lonely.
- I certainly felt disconnected from the rest of my team.
- I was disconnected from the rest of my team - there were always important conversations I would've liked to been part of.
[edited for getting the bullets formatted]
> I've got good headphones
You may not have issues with open-office plans, but you did find a way to add the walls back.
This mostly applies to jobs which involve a higher degree of complexity, like programming.. but as you are on HN, I assume you meant something IT related.
I've met plenty of people that like to act like they're some super elite programmer that is juggling complex class structures and debugging them inside their head. In my experience so far, those people are full of shit and make overly complex spaghetti code. Idk though, maybe this is different between web development and something like game engine development.
http://en.wikipedia.org/wiki/Flow_%28psychology%29
FWIW while I think interruptions are a real problem, I have certainly encountered some people who I think demand more quiet than is really needed. But some of this may just come down to differences in how tolerant people are of noise, or perhaps how stressed they are.
http://www.northwestern.edu/newscenter/stories/2015/03/creat...
Adults have it; amphetamines help but don't cure it; the ADA requires reasonable accommodations for it.
Where I work there's a group chat for low priority questions you can check whenever you're not totally immersed in something. It's also anonymous chat so nobody is embarrassed to ask "stupid" questions.
That's awesome!
(Though, I guess, with a small enough team it probably isn't all that anonymous)
Ever wonder why that someone then complains about distractions and lack of productivity? This is why
For me, it's almost so bad that I don't bother trying to get in the zone/flow, because I know I'm going to be interrupted.
Lots of people are commenting to your post and others about the inefficiency of frequently interrupted work, but I actually think the efficiency argument runs the other way. let's say I'm working on some problem and I run into some really obscure issue coming up from a system I've never dealt with. I might kill 10 minutes of your productivity by asking you, but I could also easily kill 30 minutes digging into stack traces on various servers I've never heard of before looking for an answer you know off the top of your head.
Also, usually in a fairly close-knit team most people know who is going to be irritated by the interruption and who isn't, and they go to the folks who are friendly about being interrupted. I get these theoretical concerns about open office spaces but in real life I just don't encounter them that often.
- The team has a good culture of respecting....
For instance, in my office, there is someone here who will literally speak to their team members despite having an ongoing LYNC chat open with them, in a voice that you can hear from a football field away (opposed to a whisper/quiet/ "inside voice").
I think respect and consideration could go a LONG way to making open office plans tolerable for most.
Knowledge sharing by osmosis was a very real phenomenon there. I can't count the number of times I discovered interesting new technologies or was enlightened on some architecture tradeoffs thanks to a discussion taking place around me.
I realize that the whole experience around open-offices is very YMMV but I feel it can have many tangible and valuable benefits for the right kinds of people (i.e. those who are not too easily distracted and/or can easily context-switch back and forth).
"As a rule of thumb, you should only use MP3 devices at levels up to 60% of maximum volume for a total of 60 minutes a day" (http://www.osteopathic.org/osteopathic-health/about-your-hea...).
Out of curiosity, is there a peer-reviewed study on this? Moderation sounds good, but I'd like to understand at what intensity (dB) for what period could potentially constitute hearing loss.
The 60% cited in that link seems to be based on a 120dB max (so: 72dB), which is a fair bit higher than most consumer devices (with consumer headphone impedance levels). Some cursory research shows that it's a bit closer to 103 - 109dB for an iPhone (and similar devices), which puts it around 63dB.
For context, I use canalphones for the isolation, and so I can keep the output down. I also understand that these aren't viable and/or comfortable for everyone.
TL;DR: How long can someone sustain ~63dB without potential hearing damage?
Some useful (official) tables and charts: https://www.osha.gov/dts/osta/otm/noise/standards_more.html https://www.osha.gov/dts/osta/otm/new_noise/images/fig3.gif
And this has a nice listing of levels/safe listening times: http://www.dangerousdecibels.org/education/information-cente...
Also "60% of maximum volume" is a poor metric due to differences in headphone impedance and sensitivity.
If 60% means 0.6 times some maximum number of decibels, then this is not scale invariant. For instance 60 dB versus a maximum of 100 dB ("60%") is not the same power ratio as 30 dB versus 50 dB (also "60%"). In the dB scale, we subtract to show ratios; we don't indicate ratios between decibels. For instance, 10% of the maximum volume (in terms of power) means 10 dB below.
If 60% just means 60% of the volume position between 0 and MAX, that is also hokey because it depends on the actual taper curve. Analog volume pots use various ad-hoc "audio taper" curves. Digital volumes tend to be better: the steps based on dB increments. The increments are not always the same across the volume control range.
The maximum volume varies between amplifiers and headphones. What does that even mean? Maximum level at what percentage THD? Many sound systems can go a notch louder, if you accept greater distortion, and some people listen to rock that way.
A sensible starting guideline might be, say, "at least 20 dB below the maximum output intensity which 'typical' headphones can reproduce without distortion". That is still hokey in a number of ways, an important one of which is that the user usually has no interface for dialing in a 20 dB drop; such markings are only found on the controls of some pro audio gear.
(I'm partial to sennheiser, but just get something nice)
http://www.amazon.com/Audio-Technica-ATH-A900X-Monitor-Headp...
Oh, it has been mentioned. https://news.ycombinator.com/item?id=5078314 I remember :).
https://news.ycombinator.com/item?id=5078026 http://theheadphonesrule.com/
I'm toying with the idea of recreating a library-like environment that is super quiet and zen-like, but not limited in number of people. Go there for flow-type work. On the flip side, have a cafe-like environment for animated collaborations. Of course, there needs to be a great insulation between the 2, perhaps have them on a different floor.
The first time I studied in a quiet library, I was blown away by my ability to sustain a flow state. The atmosphere, the minimalism and reduced distractions helped tremendously. That said, I had to keep my guard up and avoid falling into a routine of "going to the library" with the expectation of GTD automagically. To that end, I would consciously sit in different places (to mix things up) and make sure I had a To Do list which I worked through.
Edit: Another key revelation is the expectation that if someone needs to make/take a call, it was expected for that person to step OUT of the room. Remarkably, this was the status quo.
Also some people need privacy lots of the times (i.e., HR, senior management) as they discuss sensitive topics. Unless your organization is structured to be completely open, some folks are going to need a near-permanent office.
At the same time, I think visual distractions also need to be taken in consideration.
Specific things that I'm interested in:
1) How would folks relate the understanding of what is going on at the company now, and compared to before?
2) How do folks compare their effectiveness in getting things done for themselves? for others? as a group?
3) What is the one thing they would change?
When I joined NetApp it had just acquired a caching company, the original company had open plan, NetApp had offices. They kept the open plan for the caching employees for the transition and then offered anyone who wanted to move into an office or a cube that option. A number of people took them up on that offer.
In the non-open plan version the same people got less done and felt more out of touch than they had when they were open plan, but they enjoyed their work environment with an office more than they enjoyed the open plan space. From a company perspective it was clear that it was in the company's interest in doing open plan, and in the employee's interest it was better to do offices. Of course as an anecdote it provides no statistically valid data to the debate. But it left me with the questions above which, in a different experiment like Wildbit's I would love to have another data point.
My anecdotal data is that people in an open plan sling more code, but better architectural decisions that don't require as much rework down the line (and as a result smaller 'output') come from private offices.
I would love to come up with a good architectural quality test that was more quantitative. I've got a script that can go count how many times a given function or module was rewritten but running that on our code base of a few million lines of code really doesn't correlate with a 'gut' feeling of good architecture or bad architecture. For example, the Blekko crawler has essentially been rewritten from first principles three times over the 7 years we've been doing this. Generally that was not a result of "bad architecture" so much as it was unknown unknowns, basically writing a crawler that can run unattended for weeks or months is a really complex and often subtly nuanced problem. And sometimes is clear that the rewrite was just because someone didn't understand how something originally worked and so re-implemented it in a more understandable way. What I'm getting to is that code "touches" are not the quantitative measure I'm looking for.
It's actually the biggest reason why we have a chef prepared lunch. We found that it forces everyone out of their desks at one time to hang out as a team. We usually don't even talk about work, which is the way it should be.
The other assumption is that the break out rooms will become longer term project team rooms. So if a couple of people are working on a feature for a few weeks they can take over a room to focus together.
I can say for sure that being a remote team enabled us to work around a lot of these concerns. It's easy to feel isolated when you are remote, so we work hard to make sure all communication happens asynchronously or in chat, even if we are in the same room. We also use video a lot, and my hope is that we can use it even more in the new office with remote team members.
Sort of like the guy whose car broke so he is taking his tractor to work. He would tell the world how tractors are so much better and have all these great advantages compared to regular cars.
For some reason people believe the "collaboration" argument without evidence, but dismiss the evidence that productivity is lower in open offices. I suspect they are extraverts or high stimulation people that cannot grok introverts or low stimulation folks. When I was in school I could never study with music, for example; others require it at loud volume to even settle down.
Result: Everyone was unhappy
And that's really the whole point.
It's why they are popular in Japan, together with the fact that thanks to the open structure, transgressor can apologize in front of everyone at once.
Half of our team is remote. Using physical space for controlling productivity (or lack of) is not a solution. You and the rest of your team should keep each other accountable through performance metrics that are mutually agreed upon. Being a remote team for so long (15 years) has forced us to find alternatives to traditional office babysitting and it has helped us design our new office as well.
I never noticed anyone caring that someone came in early, arrived late, left early or left late or taking a long lunch. They cared they got their work done in a reasonable amount of time.
That certainly the attitude I saw at Google. I assume it's the same at Facebook. Work wherever (office, home, coffee shop) and whenever you want. VC in for meetings if you're not on site. Get your work done.
Note: I also worked in Japan for 7 years. Both Japanese companies, both there was at most one other non Japanese employee in my division. One had strict hours, the other not. Of course the strict one effectively took roll call. But the other didn't. The strict one started with team cubes, switched to open office when they moved to a smaller place where there wasn't enough room for partitions. The other had personal-ish cubes. Most people had their own cube with partitions high enough you couldn't see over them when sitting down but could when standing up. A few at the ends of aisle shared a space for 3.
that is what basically everybody is doing around me. In that respect open office has helped to get rid of any sense of shame we may have been able to feel before. "Mass presence effect reinforced behavior" i'd say.
If there is an employee an employee who is constantly leaving early, coming in late, and not working, management needs to deal with this directly. An open office plan is not a good replacement for good management.
Any office or managers who feel the need to make note of when I enter and exit is a great red flag for the rest of the organization. Just a personal observation!
If anything, this prevents one person from being singled out because it's difficult to get mad at one person for slacking when everyone is slacking to a similar degree.
At least, that's my experience being in a 4 person cubicle. It built trust between coworkers and against the company.
If the boss is not there, then it's not quite the open office; it's a slightly strawman version of the open office which doesn't quite include everyone. Even so, the boss can come out peek at the workspace at any time and tell at a glance who is there and who isn't, and who is doing what.
Looks at HN karma. Yeah, if only they did that.
I guess a lot depends on company culture, but I'm yet to work in an open-office space where I could actually focus on anything harder than the most mechanical coding tasks.
Well said. Some people like to blame it on founders too cheap to buy decent furniture, but rearranging people every few months is time consuming and disruptive (It can also kill morale when someone loses a window or gets stuck under a vent.)
It's not even about open or not. But I can only be really productive if I'm completely alone.
Yes, for programming at least, concentration and productivity is better when you move from an open office to something slightly more insular... But you're still thinking within a traditional office paradigm. If your goal is truly increased productivity, let's talk about remote flexibility.
Yet... for stake holders, it just seems wrong to have your workers be too independent. "Company culture" and "we're all in it together" is great for company owners, but at it's core it's really just a now widely normalized practice to get workers to invest even more energy and mindshare.
I can only imagine an increased openness to remote work setups (for geographically local workers) in the coming years. Makes too much sense.
Which is better for your career? Hint: it's not remote.
Really hard problem that needs to be solved real fast? Call me on my cell phone if a fire breaks out, otherwise see ya tomorrow.
If I had a private room (or even a cube on a quiet floor) I'd stay at the office instead.
It was interesting to see how different coworkers liked to communicate. Some people strongly preferred phone calls, while I strongly preferred IM - I can carry on a conversation or two via text without fully dropping out of my flow, while engaging speech/hearing seems to require a full context transition.
From actual experience having been there, we saved money by simply not doing that. We did move cubes and offices on a fairly regular basis such that someone "tortured" by the air conditioner wouldn't suffer for long and likewise I only had a nice office for about a year.
WRT colocated teams, a long time ago I sat kitty corner to a guy who programmed FPGA based ISDN channel PBX cards for a living, and I should care why, exactly, as long as we were good neighbors to each other, which we were? I sat next to the guy who was in charge of OSPF internal routing despite my being a BGP guy and my BGP partner in crime sat fifty feet away, and it didn't really matter much because some of my BGP peers sat two thousand miles away? When I got a call in the middle of the night to fix something, I fixed it, not drive to one of my teammates houses so I can sit next to them before starting work.
Architecture decisions can help. All cubes must have a visitor chair preferably 2 or 3, although people "illegally" stacked manuals or threw coats on their chairs if they wanted to keep people out. A bit of work is necessary to stop people from sneaking up on you by orienting desks facing the cube entrance.
One of the most shocking things I've seen in open plan offices is the visual clutter. Its like working in the dirtiest most distracting looking house I've ever seen. I have junk laying around on my desk, but most open plans look like a tornado struck a Walmart. What a heap.
I got to visit Autodesk in Northern California in mid 1990s. In that office each employee was in an enclosed office.
Down a long hallway were rows of offices with enough room for a desk/chair and 2 chairs in front of the desk for visitors. Wall facing the hallway was glass. Wall between rooms were actual dry walls.
I remember some friends who were even the most junior level engineers were given an office. When they were working late, they could just put down a sleeping bag under their desk and take a nap. None of the shared napping room non-sense.
I think that arrangement is the best.
That said, what annoys me the most about the open plan office is also having a boss who cares a lot about what's currently on my screen. (e.g., hacker news rather than vim). I hate having the instinct to cmd-tab every time I see someone walking by. (even if I'm already doing work!)
My old dorm-room had this layout: http://i.imgur.com/FB3PGYT.jpg which would work great for a two person office.
I do like the idea with having showers available.
I wouldn't be surprised at all if having such facilities, and allowing for naps, would have a measurably positive effect on the work environment. Assuming of course that the people who work there are willing to and wanting to use such facilities.
Some people might be very active and healthy in their free time, but I wonder if allowing for more healthy habits during those 8 hour stretches of often very sedentary work would even be a boost for those people.
I would guess, also, that has outer wear (e.g. nylon shell jacket) that he doesn't bother washing at all. Am I right?
I had to lay down a protective force field of Lysol when he came back from the "box".
that is truely a hundreds millions (or even few billions) sales pitch (cue architectural sales pitch from "How i met your mother"). Now even i understand that those walls/etc... happened to be just "limitations" which our internal caveman was hiding inside.
Beehive.
"The challenge was to create the most collaborative, creative environment possible. The result? A desk that we could all share, literally--4,400 square feet of undulating, unbroken awesomeness to keep people and ideas flowing."
You can't make this shit up.
http://www.inc.com/christine-lagorio/meet-the-barbarian-supe...
while progress can't be stopped, it is possibly to capitalize on it - i think you should file a patent together with the other guy who mentioned beehive :)
It can't get worse until every typo is put onto a Jumbotron as you make it, and your progress is reviewed at daily standups where everyone is dressed except you
- I really like collaborative spaces/pods except when I'm trying to really hammer on work, independently. I don't like being isolated. If I come into the office at 7-or-8am before everyone arrives, those two hours will outstrip the rest of my day (except after 6pm). Noise, mid-afternoon, is especially bad and distracting. Some people in my group have taken to occupying entire meeting rooms just so they can focus in quiet.
- I am most productive at University in quiet spaces. The library, mostly. BUT, in an environment where I'm working with other people, bouncing ideas, etc, it's not a productive environment because libraries demand QUIET. My second favorite place to work is in one of our "group study" rooms. They're small, private rooms with space for up to 6 people. Whiteboards and TVs, outlets, etc are provided.
I think separating people into these rooms would be my ideal configuration. I'm not sure that I'd necessarily partition people officially (except in cases where people want an office to e.g. store belongings), just give people a notebook and say sit where you want. I can therefore sit in whichever room I want, and multiple people who demand quiet working conditions are able to share a space. This combines the nicety of social interaction, being able to discuss with someone about work-or-non-work related things, with long stretches of quiet, uninterrupted worktime.
The ideal office (or at least one I'd like to try working in) would be where there are many offices small offices that represent these group study rooms. Natural light (with blinds), comfy chairs, TVs for presenting, whiteboards, climate control, etc. Take one room and copy/paste it. Add some open areas as well, for people who prefer to work in those spaces. Destress/common areas.
We believe that when you come to work you should make the best of those hours so you can spend the evening with your family, friends, or just relaxing.
Hopefully, if there are enough productivity gains and as the startup transitions out of super-lean mode, you could represent that increased productivity with a shortened work day/week!
I think many companies use open office formats out of principle. Misguided principles, but principles nonetheless.
A layout I've worked in, and that I'd happily work in again, is this:
+---------+---------+
| | |
| | |
+-----+ -+- +-----+
| |
| | | |
+-----+ +-----+
| | | |
| |
+-----+ +-----+
| | | |
| |
+-------| |-------+
The bottom and middle side rooms are private offices. They should have doors that close and be reasonably insulated from sound, so that a worker can work without disturbance when they want to. Ideally, the wall wall facing the central area should have a big window (with drapes or blinds!) so that the person in the office can see if anything interesting is going on in the central area. Each office should have its own light switch capable of turning off all lights in that office.
The top two rooms can be bigger offices, or conference rooms, or break rooms for breaks that might be too noisy in the central area. The gap in the bottom wall is the connection to the hallway.
With this environment, you can easily work in private, no distraction mode (go into your office, close the door, and close the blinds, and you can even play some music on speakers without disturbing others if that helps you work), or in full social mode (take your laptop to the middle area), or in between (work in your office, but leave the door and window open, so you can keep an ear and eye on what's going on in the social area.
Note that if you have two groups working on different things, but that have a manager or senior engineer working with both, you can extend this concept and put the two groups side by side, with a large office (or more) in the middle connecting to both:
+---------+---------+---------+---------+
| | | | |
| | | | |
+-----+ -+- +-----+-----+ -+- +-----+
| | |
| | | | | | |
+-----+ +-----+-----+ +-----+
| | | | |
| | |
+-----+ +-----+-----+ +-----+
| | | | | | |
| | |
+-------| |-------+-------| |-------+A lot of the time, people can't get away to hear themselves think, and it stresses them out more than they realize. Imagine never feeling like you're as focused and productive as you're used to being and the impact that might take on your self appraisal; it's like never being able to get comfortable in bed, night after night. Using the plugs helps people get away mentally and maintain their focus when they can't separate themselves from their environment.
Also, the people who might more casually chatter at you stop when they notice you've got them in without you having to decide whether to be polite and acknowledge them, which besides sparing your focus might help the ones who have a harder time controlling their socializing reign theirs in.
One recommendation to wildbit, though - that staircase with the neon-green diagonal lines may make some people nauseous as they walk down it. We have one similar to it in one of our offices and some employees have to avoid it because they get vertigo walking down it.
Around here I have visited one company where it seemed like management where actually out in the open and I think I know one major branch office at one major telco that also does (or did? I haven't heard anything the last few years.)
Take an hour or so and get a hearing check done by an audiologist. This will give you a baseline. If there's nothing wrong, great, you can check again in 5-10 years. However, if there's problems you can alter your behaviour and hopefully preserve your hearing. And if "oh s*ht" results from such a test, you will find out that hearing aids cost 10-15x great earphones/headphones and insurance tends not to cover them. In North America at least.
However, are we, the fans of private offices, claiming that our corporate overlords are not only greedy, but misguided? That in fact individual offices would be better both for their bottom line and our happiness as programmers?
I ask this only because I find myself increasingly suspect of all arguments which end by declaring that our stance "would be better for them anyway!" I call it arguing on both hands. As in, "on the one hand, this circumstance favors my position, but on the other hand so does this one!"
Cost. Flexibility is a very, very distant second.
Anyway - those offices look amazing. Great job.
The drop in center had three separate rooms. Two were loud rooms, with phones at each desk. One was a smaller "quiet room", with no phones, intended for people who do quiet, focused work. I suppose there was some "back visibility", since the workstations were set up on round (might have been hexagonal) tables, so it would have depended on where you sit. Another big factor in the success of a quiet room was an office manager who absolutely enforced the noise rules, since a lot of people who would push the limits if allowed (when spaces in the loud rooms were all taken)
There were also a few offices/small conf rooms for meetings and so forth.
I think it worked, but that's probably because it didn't have the same goals of "collaboration" or "openness" that modern open office proponents often claim to provide. Most importantly, the open office and quiet room was clearly not intended to keep an eye on workers in any way. It was a drop in center in the first place, so while you might see the same people around, nobody was really monitoring anyone else. Also, there was no status associated with where you sat - open seating, grab what you like. And of course, the quiet room showed a great respect for people like developers who need extended periods of quiet focus.
It's also clear that the "open office" quality of this drop in center wasn't based on flawed notions of increased communication or collaboration. At that time, developers at Sun had their own offices (well, I did, and I wasn't high on the org chart), so the drop in center was for convenience. It was set up so that people who had meetings in other spots, or who wanted to telecommute, had a place with a workstation if they needed it, and had a phone or a quiet room, depending on how they needed to work that day. The idea that the office would be a place for constant open communication wasn't part of the plan - in fact, the quiet room ensured that this sort of thing didn't constantly distract the developers.
Kind of depressing, now that I think about it, that the only time in my life I had my own office was my first job. The industry has really moved away from that model over the last 15 years. Facebook has moved into the old sun offices… did they tear them all out and replace them with open offices?
I also had shared offices quite a few times, and I enjoyed those (actually, I liked them more than my own office in many ways, provided I shared with another developer). Cubicles and open offices are pretty horrible, but I think that if we had a "quiet room" approach to them, they might not be so bad.
I think that 90 percent of companies that are now using them are doing so for more respectable reasons (either they believe that "collaboration" shit, or they are aggressively cutting costs... which is unpleasant but not evil) but, for many in the past, one of the motivations in the move toward less workplace privacy over the past 20 years was... to push out older programmers.
It doesn't make sense when there's a lot to do, because the older programmers are often the most efficient in terms of value per salary (i.e. they get 5 times as much done, but only cost 2x as much). You want old programmers when you have too much work to do. But when you have a slowdown and don't need high-end work, those older programmers will be seen as (expensive) excess capacity. If you're also looking to shave costs, moving to a crappier office space kills two birds with one stone.
All of that said, I doubt that open-offices have the discriminatory effect now that the entire tech industry uses them. They used to push out older developers, but now, the people who absolutely can't stand them (or whose health can't handle them) have long-since left, and they're just making everyone's life hell.
Of course, these offices also discriminate against people with a large array of health problems, and they aren't exactly pleasant for women... and I wouldn't be surprised if there were a massive class action suit in the next 10 years over it. It'd be hard to prove discrimination in most cases, but I'm sure that the bigger tech companies have discussions about office layouts and age that wouldn't look good in discovery.
They're also not a great economic trade. Office space is really cheap in comparison to having your people be distracted. $40/SF per year would be high-end commercial real estate, and you only need 150-200 SF to give developers a decent layout.
Loads and loads of stuff in business runs on cargo cultism.
It is very, very hard to say what precisely causes businesses to be successful or to fail. Correlation does not equal causation, etc. As a result, you have lots of "fewer pirates, higher global temperatures" type conclusions being drawn.
The discriminatory motives you cite factor into it -- correlations get promoted to causation when doing so will reinforce one's worldview or otherwise lead to a desired outcome. Sometimes this is subconscious.
Real heavy cargo cultism kicks in when someone with "juice" (a "thought leader" -- gag) draws such a conclusion and codifies it, or when a sufficiently large number of people independently do so and a herd mentality kicks in.
Edit: Is it the smoke from all those dollar bills going up in those flames?