Its obvious on the face of it: re-usable walls and shelving are going to be more expensive than simple stick construction. You have to be re-using it constantly (reformatting office space every month or quarter) to make it pay. And then you're tanking everybody's productivity.
I think open-office is some brain virus that keeps infecting managers everywhere. We need some kind of vaccine to combat it.
The vaccine is a startup that is 10x (or even just 2x) better than everyone else because they use private offices.
Since that hypothetical business hasn't yet proven that idea, all the articles from journalists writing about "open offices bad" are just preaching to the choir.
Even the common cited reason for open offices being "saves real estate costs" is questionable. As an example, look at Mark Zuckerberg's old Harvard photos when building Facebook. Look specifically at the 8th and 16th photos.
See how everybody is literally in an "open office" crowded around a kitchen table?
In Mark's mind, that collaboration "works" for him and helped make Facebook successful. Therefore, it should also work for future hires. This is why cash-rich Facebook that has money to build private offices equal to lawyers' suites eschews that and opts to build an open plan instead. The new 2015 headquarters is expansion of that "2005 Harvard open office" on a grander scale.
Mark Z works still works in that open warehouse concept instead of a private suite.
I see very little commentary from HN that directly deals with executives who really believe in their hearts it's a superior way to work.
 deep link: https://www.youtube.com/watch?v=l--zev_37QA&feature=youtu.be...
Fog Creek has(had) 3 major products:
- FogBugz: profitable but less of a success than Atlassian
- Trello: not profitable (sold to Atlassian)
- Stackoverflow/StackExchange : not profitable yet 
According to the open-office "distractions/interuptions" theory of killing productivity, the Atlassian programmers should have been severely handicapped and as a result JIRA should have evolved at a snails pace. Instead, the opposite happened and Atlassian JIRA released more features than FogBugz. Both FogBugz and Trello lost to Atlassian.
That Joel Spolsky post about private offices gets repeatedly cited in threads about its benefits but I recommend people not mention it. It undermines their point. It's ineffective at convincing executives. However, it's very effective at making other programmers reading it nod in agreement (aka "preaching to the choir").
Don't link ineffective articles devoid of business evidence that happens to confirm your desires. Instead, study the way some executives actually think. Too many programmers dismiss companies' rationale for open offices merely as "saves square footage costs" or "it's a way to spy on employees because of distrust". Yes, some of that may be true but others also have different reasons. (Take a look at the Mark Zuckerberg video I linked and listen to what he's saying about his desk in the open floor plan. Is he trying to recreate that elbow-to-elbow collaboration he had at the Harvard kitchen table or is he just trying to spy on people? Would that Joel Spolsky article convince Mark Z to build private offices? No? Why not?)
At Trello everyone is either remote or has a private office. I'm not sure how to prove that they're 2x (I'd say it's too vague to be provable), but they've managed to do well without taking large amounts of investment, which speaks well of their productivity.
I think you're inadvertently undermining your point. Trello wasn't profitable. They were "cash-flow break even" which is also another way of saying they still had not earned enough "free cash flow" to pay back their past internal investments that got them where they currently are.
Trello has private offices.
Atlassian has open offices. They are also profitable.
Atlassian was the one who bought Trello. Trello didn't buy Atlassian. Trello did not perform 2x better than Atlassian JIRA. (E.g. the ideal narrative would have been, "because Trello programmers have less interruptions than Atlassian programmers, their productivity was proven to be 2x superior and they made Atlassian JIRA obsolete.")
If you want to change the hearts & minds of people like Mark Z, the Trello example is not a case study to use.
Instead of using Trello, you could compare Atlassian against Fog Creek. Fog Creek was founded two years before Atlassian, and both companies are more or less in the same space - they make productivity tools for developers. And yet it's hard not to look at Atlassian as the more successful business so far.
So, recognition of need for deep concentration and focus, but in an environment totally inimical to those.
Only if you "decide" to be informed immediately when an email/message comes.
With IM you can turn it off if you need to focus, and turn it on when you are doing busy work anyway.
As a person that has worked in both open and closed floor plans, ill just say this, I miss working from home and cant wait to go back.
5-10 years after the initial deployment, the same models are no longer available and the new cubicle system isn't compatible with the old one so they're stuck with the now obsolete setup and are forced to either work with it or scrap it and start over.
Cubicles or open plan is by far the cheapest way to set up an office since you only have one room that you need to set up. And that's the main reason people use it, it's just cheaper. If you look at flexible working spaces, desks in closed rooms often cost twice of what you pay for a desk in an open floor layout. I don't think there's a way to set up closed offices at even close the price of open offices.
And I must admit as much as I don't like open offices, sitting in a tiny private office of the size you have to yourself in an open office would probably lead to anxiety.
You can always double-up in a 10X12 office if you grow. And until then, its very nice. And cheap.
Flex time with small group office is nearly ideal. I can spend hours a day with the door closed concentrating or meeting, or hours a day working as a team, it just seems ideal.
Its like the difference between college dorm life with a roommate or two, vs military barracks grid array of 50 beds packed together.
The problem is someone noticing that and then sticking 4 people in your office.
The more polite - "We're 100% contracted out right now" is somewhat less effective - they come back a month or so later.
I always got the feeling it came from some C-level exec giving the thumbs up to the Facilities exec's hateful (but ignorant!) Powerpoint about openness and collaboration and TCO and the fungibility of talent.
"Well of course we value openness and collaboration just as much as Facebook does, and they're open-plan!"
At my previous cube farm employer, we had maybe 75 cubes, a giant 30x30 cafeteria/lunch/meeting room, oh 16 tables to eat lunch at least, a large conference room we literally called the large conference room of 20x20 and a small conf room we called the small conference room of 10x10 and there was an engineering team meeting operations room (really a lockable storage room / lab) that was 20x20. Because coats and boots "can't be stored in cubes" although we did anyway, there was a row of 50 feet by 3 feet of coat closet that was basically unused. That's a lot of square footage allocated to no individual therefore "saved" but offices would result in 2000 or so sq ft of shared space being eliminated. Now figure a 10x10 office shared by two people, thats 40 people's private offices just being wasted in the shared space required by cubicle life. So of the 75 people in that office 40 are in the new offices and 35 are distributed in the space formerly occupied by cubes. Certainly cube walls are slightly thinner than private office walls but the space savings won't be a factor of two. Definitely the employer was throwing away a considerable amount of expensive rent by using cubes and meeting rooms instead of private offices. If they junked the cubes and went private shared office they would have still had extra leftover space maybe for fancier larger offices or some people could have solitary private offices or maybe some "neutral ground" meeting rooms.
Yes, if they provide special "collaboration rooms", that's going to add to the cost. But if they don't, then it's not a factor. If your employer just gives you one big open room with a bunch of tables, and that's it, that really doesn't cost much. And there's a bunch of employers these days that do exactly this.
(As for coats and boots, you can put your coat on the back of your chair. Or drop it on the floor under your desk. Yeah, it sucks, but again there's plenty of employers that treat engineers this way these days.)
There wouldn't be need for that many rooms if there were proper offices.
But yes, if they had proper offices, they wouldn't need many of those meeting rooms, only some larger ones for meetings that are too big for the offices (more than 3-4 people perhaps).
Metaphorically speaking, many of our peers believe that this vaccine causes autism. It occasionally does cause illness more severe than those vaccinated against.
It's collective bargaining through a cartel of skilled laborers. Unions.
But like Brundlefly, those infected managers don't believe they have a disease. They're not getting worse, they're getting better (as they turn into monsters).
We don't need unions - we need developers who strongly refuse to work in open offices. Since there is a shortage of developers, this should suffice. The large problem is that too many developers are willing to compromise.
...but we don't need unions?
The ethical, nonviolent way to reject bad pay or working conditions is to quit, accepting that the employer might find someone else. That's not what unions do.
Your adversary is not going to be ethical and nonviolent.
Unions resolve the prisoner's dilemma in favor of the prisoners. The game is set up like this:
In each trial, 3 players distribute $300.
A and B vote on whether E gets $100 or $150.
E can cast a tie-breaking vote.
E decides how to distribute the remainder to A and B.
| A gets | B gets | E gets |
| $100 | $100 | $100 | A $100, B $100
| $150 | $ 0 | $150 | A $150, B $100
| $ 0 | $150 | $150 | A $100, B $150
| $ 75 | $ 75 | $150 | A $150, B $150
I mean, I get why some people don't like them, but let's not pretend that the sentiment is universal.