Science, Superstition, and Open Plan Offices | Hacker News Reader

Science, Superstition, and Open Plan Offices

dandreamsofcoding.com

51 pointsbratfarrar58 comments

I interviewed with a company and during the "cultural fit" interview with the VP of Engineering, I asked why they used an open-plan office. I mentioned that there seemed to be a lot of concern in the literature and online about developer productivity in such offices. He said he "wasn't familiar" with the discussion, but he didn't think it was a problem.

I was offered a contract-to-hire situation, instead of the employment I was seeking, and was explicitly told that the reason was because I seemed to be concerned about the open office. So be careful about bringing up "science" with people who aren't interested in it.

For someone to be a VP of Engineering and to not even have heard of the debate around office layout and developer productivity is a bit alarming.

Presumably you didn't take up their offer?

Interviews are tough. There's very little time to get a feel for whether a person would be a good fit for an organization - and often, you have to make a judgment: was this comment just off-the-cuff curiosity, or does the fact that the person brought up the open-plan indicate that he or she would have problems with the way that the company works?

In interviews, the open-plan office is non-negotiable. It's not like you're going to convince them in an interview to change their minds and build closed offices. I can imagine a lot of interviewers going "well, if he or she brought it up and has actually studied literature on it, then it's probably something that bothers them, and if it's something that bothers them, we might have a problem down the road... maybe it's better to test the water than to offer full employment."

To be clear, this is a problem with the interview process, not with you (or the person hiring). There's just not enough time to actually understand a person, so interviewers are forced to make guesses and go with their hunches. And because the cost of hiring the wrong person is high, interviewers will often look for any small thing to refuse a hire.

Fit interviews are hard for engineers. Technical interviews are about getting the right answer. Fit interviews are about whether you can tell a convincing lie ("of course I like the working under tight deadlines to changing requirements with minimal support from the business.") It's a tough context switch: from zero-in-on-right-answer to socially-acceptable-bald-faced-lie.

I half-agree with you, because there are some companies that use "fit" to determine how much they can use you as a "resource" rather than a human being.

Where I disagree is whether one has put themselves in a situation where they are desperate for a position at a company where they have to lie to pass the fit interview stage. I'd rather be honest and lose than lie, win and ultimately impugn my integrity.

It is very refreshing when you already have a job and are looking for something new, then you can sit back and take a relaxed approach to finding a good company. Once one is found and you are interviewing with them (and interviewing them, as a previous commenter mentioned), then you can ask all the tough questions. If you don't like what you hear, vote with your feet and walk away.

Now, this isn't always the situation people are in, so YMMV but I throw it out there as a piece of advice for anyone interviewing companies.

I fully agree with you... and yet we disagree on a matter about the outside world.

My experience is that good companies are about 5%, at least weighted by jobs that are on the market. (There could be more good companies that are small or rarely have to hire.) I might be off and it might be higher, but not by much. They're rare, and they're selective, and they tend to grow slowly.

Most people are going to have to swing from one crappy company to another until they find something good. Once you have something legit, you definitely want to hold on to it for a good few years.

Also, on a practical note, there are good teams within crappy companies that are still worth working for. I wouldn't advise, say, most large tech companies or investment banks in general, but most of those huge companies have teams that are really good and make it worth taking the job.

I'd rather be honest and lose than lie, win and ultimately impugn my integrity.

I used to feel this way, but ultimately I'm of more use to the world if I'm occasionally willing to part with honesty. If I see someone in the grocery store who's 400 pounds, shouting "You're obese and if you don't stop eating that crap, you're gonna die!" does no one any favors. It's "honest", but worthlessly so.

I often wonder why companies offer contract-to-hire positions when so many regular employees are at-will anyway. I am especially puzzled as to why a company would offer it as an alternative to both full-time regular employment and a fixed-term contract.

It seems that mixing the two removes all benefit to the worker while reserving all benefits to the company. So why would anyone do it if they had a choice?

I could also see this as the company giving the worker an easy out based on the concerns they expressed about the open office - there is no worry about waiting for the "right time" to quit if you're not happy with the new position when the "right time" comes in the form of the end of a contract period before full employment.

You also have to consider the opposite side of the coin- you should be interviewing them as much as they are you. The VP's responses should have been a red flag to you. Not just because of the fact that he apparently hadn't heard of the debate over open floor plans over the decades, but mostly due to the fact he seemingly punished you for speaking out. Typically, not fun people to work for.

And yes I'm fully aware that sometimes you're in a positions of "I need a job and can't be too picky". I hope you were in a better position than that and had the opportunity to turn them down. If not, I'd be curious how it turned out.

On the other hand, studies also show that "radical collocation" can increase software development productivity. This study from the University of Michigan claims that productivity was doubled when developers worked together in specially designed "war rooms" that allowed for both spontaneous meetings and moments of solitude: http://www.researchgate.net/publication/220879336_How_does_r...

This quote summarizes the results well:

"Although the teammates were not looking forward to working in close quarters, over time they realized the benefits of having people at hand, both for coordination, problem solving and learning. They adapted to the distractions of radical collocation, both by removing themselves to nearby hotelling areas when they needed privacy, and by zoning out, made possible because of the distance between people in the larger rooms."

It feels productive to enter a flow-like state for several hours debugging a problem, but it's probably more productive if a teammate notices and provides the solution in minutes. Nevertheless, stretches of uninterrupted focus are often essential.

I think that's one thing that gets missed in these discussions: software development is different from, say, accounting or administration or whatever.

I get huge value out of being right beside the people that can answer the questions I have, and I know that they get value being near me. Putting walls between people and enforcing no-interruption work inhibits communication and makes it much easier to get blocked.

At our office, I actively encourage every new hire to interrupt me and ask a question when they get stuck. (I actually get kind of annoyed when they don't). Lots of people, especially more junior devs, will sit down at a problem and really get lost for a day trying to figure it all out themselves. I'd way rather be interrupted for a 15 minute explanation of some bit of code than have them lose a day.

> I get huge value out of being right beside the people that can answer the questions I have

As someone who -- in my workplace -- is more often than not on the receiving end of questions or calls for help, I find it extremely distracting. There have been entire days where I wasn't able to focus on my own work because everyone else had questions to answer, or bugs that they needed help working through. I can't fault them for asking for help, but I definitely get frustrated at the situation, because sometimes at the end of the day I end up feeling like I haven't accomplished anything.

Oh, so do I (and I'm more often on the receiving end, too). No question, there's a cost to being interrupted. Personally, I have to actively push back against my own displeasure at being interrupted.

At the end of the day, though, the value those interruptions bring to the project make them worthwhile. Problems are avoided before they become problems. Knowledge is transferred without forcing someone to learn it from the ground up themselves. Time is saved (maybe not my time, but project-time).

(Obviously, if you're consistently being interrupted with low-value, easy-to-answer questions, that's a bit different... there comes a point where you say "ok, spend a bit more time, and if you can't get it then I'll work more with you")

Then the question is whether team productivity has gone up, even though your own personal productivity may have gone down? You probably accomplished a lot being helpful, though I know that doesn't help how you feel. And if your performance is tied to personal productivity, then that is a real bummer.

The "nearby hotelling areas" thing is okay when you do everything on a laptop, but as developers we tend to have all our state on a desktop machine driving 2 or 3 large monitors. We can't just pick it up and go into a "focus room", as they're called where I work.

My own opinion is that it should be the other way around. In the open space, library rules apply: brief whispered conversations only. For extended discussions, go into a "focus room".

That, I think, could work.

I was really sympathetic to the distractions caused by an open plan office. I saw the statistics that people get sick more often. I knew it would be annoying to have it potentially be really loud (or have everyone need to be quiet to keep it not loud). My company's new office is going to be open plan, though. Why? Because an office costs about $15k to build. We're already spending six figures on construction (ha ha biotech) and there was just no economically reasonable way around an open plan office.

I think the real reason many offices are open plan now is simply that it's much, much cheaper. The rest are kind of just-so stories, for or against. Interior designers will come up with these justifications when prompted but really I think it's that the people who actually get to make the decisions look at the general contractor estimates and say, "okay open plan it is."

Let's assume my office employs 100 people, at around $100,000 per year each. If a $15k office would make us 10% more efficient, then it would pay for itself in 18 months.

Open plan offices are penny-wise and pound-foolish.

That's a $1.5M investment up front. Maybe it will make you 10% more efficient, but I have to look at all of the different things I could spend $1.5M on. Hopefully there's something in there for a growing company with an expected IRR of more than 10% over 18 months.

In parallel, how many months of runway do you have? I'm guessing it's not 18, much less the 36+ you'd want to really see those benefits pay off even if it were the most efficient use of the capital.

Yet iF you want to save time then the only critical resource is your employees. 10% is an extremely low estimate of the disruption open-plan offices engender. In some cases they can completely disrupt some employees, resulting in 0 or negative return.

I'm talking about a 100 person office, not some start-up living off VC. At 100 people, you're going to be spending $10m on salaries alone each year. $1.5m isn't nothing, but it's close. And for a company with 100 employees, a 10% gain for that price is worth every penny.

Mind you, the 10% figure is totally made up. Maybe it's 50%, maybe it's 0.5%. But if I had to guess a number, I'd call it 10% on average.

Am curious whether you will work in the open-plan office. In other words, when the new office is ready, will you get a room with a door that closes?

Maybe we need to re-brand the "open plan" to something else. "Open" is something many tech businesses want to be, especially internally. So an "open" layout helps people be "open" in other desirable ways, right?

Well, not really. It's an unfortunate decision-scrambling word-collision.

It deserve a name that's still descriptive, but of a loud, crowded place where it's hard to get work done. "Stadium plan"? "Nursery plan"? "Transit hall plan"?

Awful Plan? Noisy Office Plan? Distract-o-matic?

How about boiler room? Has a lot of negative connotations: https://en.wikipedia.org/wiki/Boiler_room_(business)

Exposed office plan. Cheap office plan.

I really like Cheap Office plan actually. For all the ra-ra Kumbayah BS around open office plans, the fact that they are cheaper in the short run is why Cheap Office PLans (tm) are popular with management.

Hmm. Unfortunately 'cheap' also has some positive connotations in business – and it doesn't necessarily imply productivity-damaging. "It's both 'cheap' and 'open'? Sign us up!"

But breaking down 'cheap' into more-explicit descriptions of missing-amenities could work. The "infantry barracks plan". The "shantytown plan". The "disaster shelter plan". The "resettlement camp plan". The "soup kitchen plan".

We just call our largest openspace the Chinese Sweatshop. Even managers use this name.

I work in an open office environment and I spend a lot of my day positioning myself so that I'm not in direct line of sight with anyone else's eyes. I hate that I can't ponder to myself without risk of accidentally making eye contact with someone.

What makes it worse is that if I sit in an upright (comfortable and ergonomic) position I am staring directly at someone. I hate it.

As a result my posture and productivity noticeably decline as the day goes on.

Interesting. Never even considered the eye contact thing. Ours is open plan but you're always looking at a partition or out the window. That makes it a bit more bearable.

Yeah, if the cube walls are low enough that you can see over them -- or if there are no cube walls -- it's just a disaster. I was in a place like that once and hated it. Eventually I found a spot where I was facing a wall; that wasn't so bad.

Here's another quote: "For every complex problem there is an answer that is clear, simple, and wrong."

My experience, anecdotal as it is, is that an open space per team is the most effective. This being science, when has the truth ever been precisely one end of the spectrum or the other? I'm naturally skeptical of pundits, advocating science, who point out that developers with their own office perform better than developer who share an open space with the call center and then claim that this demonstrates that private offices are the best solution.

The key quote from the article is "quieter and more private".

Maybe in the days when The Architect wrote a design doc and The Engineers went into their offices for six months, then a private office was a great way to get to the "Oh fuck this thing will never work" moment faster.

In contrast most teams I work with collaborate at the feature level: engineers work on the same feature, implementing different pieces together, before moving on to the next feature. We have other teams where individual developers take a feature and go off and work on it for a few days. The "individuals" are "faster" claiming work is complete, but the "teams" are faster at producing work that is accepted.

I should also add that while the team environment works for most engineers, we have some that absolutely require a private space, and we try to accommodate them. Individuals work differently and when we find someone good we try to make the environment work for them.

YMMV.

The simple answer is that these office plans are cheaper. It's a common business pitfall. The cheaper short term solution wins in spite of long term health productivity and culture advantages.

That might be true in some cases but there's an entire industry built around renting open-plan workspaces to solo founders and freelancers who freely choose to spend an extra $5000 per year and an hour's commute time each day because they feel an "office environment" surrounded by other busy professionals focuses their mind better than the serene quiet of their bedside desk. Many of them are developers.

Some workplaces don't limit it to "an open space per team", but mix multiple teams, whose distractions aren't relevant to other teams, in the same big room.

"In contrast most teams I work with collaborate at the feature level: engineers work on the same feature, implementing different pieces together, before moving on to the next feature"

Collaboration is much better with private offices. If Steve and Jim and Sally are working on a feature together then they can all take their laptops into Steves office and work on it together.

On the other hand even something as simple as pair programming in an open office can disturb dozens of other people. Where I work teams will book out meeting rooms for a few days at a time just to be able to collaborate without annoying the rest of the floor.

I suppose open plan has some advantages, like it is more hip (makes for great "we are a family" pictures), and the "team building" might make people less likely to leave their jobs. Just guessing - in any case there might be effects that offset the lower productivity.

Otoh I was floored to read today that Amazon apparently has lots of dogs in their offices because it is supposedly good for the atmosphere, lower stress levels and whatnot. I heard Google has lots of office dogs, too. I hope this is not a case of pseudoscientific superstition, because for me office dogs would be an absolute deal breaker. If most offices start adding dogs because of some shaky research results I'll be in trouble :-(

My company is co-located with another company and they encourage dogs in the office.

My question has always been: what if you want to hire someone with a dog allergy?

Worse: what if the person isn't currently local and there aren't any dogs in the office the day they happen to come in but relocate only to find they can't possibly work in that office?

Work remotely.

I was always kind of amazed about companies that do this. I'd love to bring my dog to the office, but when does it become not ok to cause physical harm to my co-workers?

While I was at google, I had (and have) a minor pet dander allergy. There were _many_ dogs regularly in the office, and perhaps I'm being overly paranoid/judgemental of the culture, but mentioning your own discomfort would have been read very negatively against you by not only the dog owners but by all the people who do buy in to that it is some part of the culture.

Whether you read that as a feature or a frustration, it comes down to that it simply becomes another line item that potential candidates may be judging you on, and you may never know about it. (it's certainly something I keep an eye out for nowadays.)

Personally, I think it's wholly inappropriate to put an employee in that situation, especially since new employees can't speak up as freely as those with a longer tenure at the organization. You can't just show up and start rocking the boat.

> I hope this is not a case of pseudoscientific superstition

There is science behind it, although accommodations should be made for those who are allergic or have phobias.

"There is plenty of research to show that this isn’t just another shaggy dog tale. Studies going back to the early 1980s support the idea that dogs—and other pets—have enormous health benefits for people. Pets have been shown to lower blood pressure, improve recovery from heart disease, and even reduce rates of asthma and allergy in children who grow up with a Fido or a Frisky in the house. Pets also improve people’s psychological well-being and self-esteem."

http://www.health.harvard.edu/blog/therapy-dog-offers-stress...

http://www.ncbi.nlm.nih.gov/pubmed/19185195

I don't doubt that pets have a positive effect. But some dog running around isn't the same as a pet. It might be beneficial for the dog's owner, but not for me (and presumably not for other people either). It's not my pet, it's somebody elses pet.

Of course it seems possible to hire only dog lovers, who would all benefit from improved mood in the dog office. I just hope that not every company adopts the rule, because some scientist claims it will universally increase mood.

A dog approaching in the street is very stressful, especially if you are in company of a small kid.

My litmus test for an open office fanboy manager is whether he too works in the open office. Usually not in my experience.

Funny that...

I prefer open spaces.

I like to talk to my team about all things tech and all things not (we have the World Cup on all day on one of our large HDTVs while we work). I also like to know who around me is available to help if I have any issues with "Blame" code. I love paired programming and reflecting my ideas off others to get a good sense of what the correct course of action should be.

When I feel like I know what to do and just need time and focus to complete something important without interruptions, I telecommute.

In other words, the best way to work is to have the ability to have it both ways.

>"Of course, scientists once thought that humans had 48 chromosomes. That partially hydrogenated vegetable oil was healthy. That egg yolks were bad for you."

Scientists still do think egg yolks are bad for you.

http://www.sciencedirect.com/science/article/pii/S0021915012...

http://scholar.google.com/scholar?hl=en&q=egg+yolk+consumpti...

Recently i've been interested in buying a noise cancelling headphones. There's the new bose quietcomfort 20i, which are claimed at least by some reviewers to almost cancel talking sound.

Assuming they work well, since they are controlled using software, it might open the way to a more "programmable" noise environment. For example you could set them to hear only the team members you want at any given time, thus giving you the best of both worlds.

I haven't tried that model, but in my experience and from what I've heard, noise cancelling headphones don't work that well on voices. They mostly cancel low frequencies, below the vocal range.

In-ear monitors, that seal the ear canal like earplugs, work much better.

I know. They are supposed to be unique in that regard according to pcmag.

You get used to it - plus unavoidable in some cases (travelling staff plus office with "hotchair" open office seating for those not travelling at the moment). I find that as long as I'm allowed to use earphones I can block things out sufficiently.

I prefer little superstition with more science

OP is a generally good post. Two minor comments.

Open plan has the advantages of being familiar, allowing more people to fit into the same space, and being more egalitarian. I.e., since offices are viewed as status symbols, providing them to developers and not to other groups could have a significant negative impact on morale among non-developers.

Open-plan seems egalitarian, but it's not. It's the opposite. If you're a manager and your boss is on another floor, you don't have to worry about what's on your screen, how often you go to the bathroom, and people coming up behind you (only interns will do that) when you're obviously not receptive. It's subordinates who end up feeling like caged animals.

When you work in an open-plan office, power relationships are shoved in your face 8 hours per day. If you're the rare sadistic middle manager who enjoys watching people squirm, you can do "the Boss Walk" all day and enjoy it. Everyone else (even decent managers) is miserable. If I were a manager, I wouldn't want this: I don't want people to feel threatened and anxious (or a need to change what's on their screen) every time I walk by their desk on the way to take a piss.

That the engineers at your company are more productive, or more in tune with the needs of the business because of the constant natural interaction with their coworkers.

I worked for an R&D think-tank during a summer internship. Everyone had an office, but at 3:00 there was "tea" and people got together for snacks and board games for an hour. They naturally ended up having this kind of conversation.

There are better ways to achieve this. Open-plan offices may populate the social graph, but they flood it with antagonistic edges. What the fuck good is that? None.

"I worked for an R&D think-tank during a summer internship. Everyone had an office, but at 3:00 there was "tea" and people got together for snacks..."

That's also the custom in many academic departments.

2014 is the year of the war on open plan offices.

Some people enjoy working in them, you know. Not everyone is interested in saying "how high" every time some grad student's study says "jump".

Peopleware was published in 1987. This body of research is not something that suddenly sprang up in 2014.

So are people not allowed to prefer open office layouts? They're not allowed to discern what works for them, because there couldn't possibly be any variables that make the results of some study irrelevant to their unique situation?

Edit: It would appear the answer is "no", no one is to be allowed make that decision. HN hates open plans more than they hate sexism.

So meet us "private office folks" half way and have enough private offices and open space for everyone. That way if you want to work in the open office section you can but the rest of us aren't forced into low productivity and constant low level anxiety simply because you enjoy it.

That sounds like a reasonable compromise, but I'll need to see some studies before trying it.

Reply on news.ycombinator.com