Fog Creek Software
Discussion Board

private office

--So now I'm envisioning the ideal development interior plan as one where programmers have private offices but they meet in labs with their buddies to get production programming done. No email, phones, or non-programmers would be allowed in the labs. All the non-programming work you do -- design, negotiation, scheduling, reading, thinking, and Tetris -- could be done in your private office.--

Initially. my thought is. Why bother with the private office? Couldn't you just say "do your other crap at home, then come into the office when you are ready to pair up and start programming?"

My second thought is. I'm not sure what the point of the lab is. If I'm already in my private office...

ghetto superstar
Wednesday, May 1, 2002

The mythical Private Office...  no thanks.

Obviously, I'm not a fan.  Nope, nadda, zilch.

There is the occaisional interruption.  Just today, one of our engineers was explaining some source control problem to another across the cube while I was deep into a random access seek issue.  It was annoying.  I printed out my code, took the pages into a conference room and figured it out.  Not a big deal.

However, for the vast majority of my work, having the constant din of peeople thinking and working is a benefit, not a hindrance.  I hear about issues that I'm not an expert on and I listen, or I hear about issues for which I'm an expert and I contribute.  You can never get that sort of cross-pollenation in offices.

At their worst, offices breed NIH (not invented here).  Why?  Because you're isolated.  Oh yeah, since http, no man is an island.  Let me ask you this:  What do Linux and MSFT have in common?  Nothing?  Hardly.  They have one thing in common: The cult of personality.  Yup, you got it.  Hero worship, plain and simple.  Working in offices elevates individual status above the overall status of the project.  Well, that's my hypothesis.  I cannot prove it, but I can provide examples.

I've been part of some great projects - and they were not at Microsoft (I did time at M$FT, but no great projects).  I was part of Thomson's DirecTV launch, the first large scale commerical deployment of MPEG systems anywhere.  We did that project in record time (under a year from start to set tops in stores), and at the beginning there was no hardware, no encoders, no nuthin'.  Not even an MPEG-2 spec (it made CD by end of the project). And guess what else: no offices.  The result was not just roduct but an IP boon for Thomson.

Another project:  Thomson's DOCSIS modem.  Thomson and Toshiba were first to recieve DOCSIS certification among a field of about a dozen competitors.  No offices.  Just pure software and hardware engineering in the flux it was meant to be in.  Pure humm.  It's how you do team development.  All questions get answered, everyone is up to speed.  You catch fish with a very big net.

The latest cool project is Video over IP.  Not rocket science (at least not what I'm doing - its all been done before), but y'know what?  It ships.  It goes out the door, its in people's homes, and its getting better every day.  That's more than anyone can ever say about MSTV.  And y'know what? NOTHING new came out of that project.  Nothing.  Because we were reinventing the wheel on our own little offices.  Pathetic.  (Do not confuse MSTV with Ultimate TV (formerly WebTV).  UTV provided the highest end satellite receiver known to consumers, but since Redmond didn't invent it, they are shutting it down).

Ingrown or outward looking.  Offices make a difference,

Nat Ersoz
Wednesday, May 1, 2002

>>> "Initially. my thought is. Why bother with the private office?" <<<

The basic flaw in your vision is doing the programming in the lab.  Common areas are for meetings.  Or for special hardware if you are doing hardware related development.

Programming is done in your office, along with all that other stuff.

Wednesday, May 1, 2002

I feel a good employee views his firm's expenses as their own.  B/c indirectly, they are.  I would NEVER blow big money on frills like fancy oak desks, so I'd be disappointed if my employer did. 

I figure, I get paid to perform a service, and I'll sit whereever they want me to.  It doesnt really matter to my end product. 

I've used cafeteria tables, shared offices, and roomy spacious luxury desks.  I've had my own phone, and I've had shared group lines.    The more spartan the office, the more money that is available for "real" things like PC's, software, training, and paychecks. 

As a result, I never put my office surroundings very high on my list of needs, measure of quality, or my measure of worth to the firm.  That's what paychecks are for.  I've always been paid well ahead of my experience curve, so I've never felt "lowly"

I never have had an office.  I don't want one either, b/c I bet it's hard to go back once you get used to that.

Wednesday, May 1, 2002

>>> "I would NEVER blow big money on frills like fancy oak desks, so I'd be disappointed if my employer did. " <<<

The people on this discussion board who want to argue against private offices for developers tend towards strawman arguments.  Has anyone ever said that developers need a fancy oak desk?  Do offices need to be "plush" as another post stated?

The best office space that I have ever occupied was an old school house converted to office space.  Nothing special there.  Offices had four walls of plain sheetrock, a door and a window.  Furniture was a generic office desk, file cabinet, computer table and small book shelf.  It was adequately quiet.  It was great!

The cost difference between a cubicle and private office amounts to about 2% of my salary.  I wonder about the thinking of a company that pays me to take phone messages on a shared phone line or to listen to my officemate crack his knuckles and drum on his desk rather then spending another 2% and letting me spend full time developing software.

>>>"I never have had an office. I don't want one either, b/c I bet it's hard to go back once you get used to that. " <<<

You win that bet.

Wednesday, May 1, 2002

Here is a hypothesis on the office vs cube vs desk thing thing:

preamble: In my last 2 jobs I have worked in 1) a shared office 2) a cube. I noticed that in the office I was more productive, but I do not think the reason for this was that it was an office, I think it had more to do with the fact that fewer people came by my office to bother me, so my hypothesees are below

item 1: that one's ability to work well is directly proportional to the amount of traffic heading into and out of your cube/office etc ... (it does not seem to matter if your area is office, or cube

item2: that the amount of traffic into and out of your cube is directly proportional to the % of incompotent people in your organization, and inversly proportional to the competence of your manager

(that is, you will get more visits: from other groups support, engineering, tech writers, QA, PM, Project managers etc) and calls from your manager.

item3: that it is not really an office or cube but your geographical location i.e. crossroads, vs out of the way cubes/offices that  affect traffic.

item4: that offices do help in the sense that you can close the door and put a "Fuck Off" sign on it, and that all things being eqaul quiet is better than loud

item5: that many of the benefits of offices, low foot traffic, quiet etc ... can be very quickly canceled out by managers who a)dole out pointless projects b)insist on people working on multiple things at once c)insist on hiring incompetant people to generate traffic in the first place d)insist on more than 1 simple status report. E) create work for political reasons (i.e.prototype this for me in C# so I can win an arguement at the next meeting)

item6: that many managers who treat workplaces like a carnival (high traffic,high noise...) are also the same ones described in item 5 (though not always)

Daniel Shchyokin
Wednesday, May 1, 2002

Mackinac has the right idea.  Offices are not very expensive overall and they are perceived as a non-cash benefit.  Delusional or not, an office still has some prestige value.

What some may be overlooking here is the indirect effect of employee morale and retention.  Turnover of bright people has great cost.  If people "nest", turnover can be reduced.  Offices facilitate that.

Really, I can't see any way one can run the numbers to show that offices aren't a good financial deal.  Not some jewel-encrusted dot-com fantasy, but a basic, modern, functional space.

Again, offices are cheap and a bargain relative to what's at stake.

Bill Carlson
Wednesday, May 1, 2002

Office space costs are quoted by square foot.
However, is this annual or monthly ?

An 8x8 cubicle has total sq. ft of 64.
At $50/sq/ft, that's $3200.

Yea, the cost must be annual, otherwise the cubible lease would be 38k annual.
I figure an office 4x the size of a cubicle...which puts it at $13k.

What's my point?  I wish I knew.

Wednesday, May 1, 2002

>>> "Office space costs are quoted by square foot. However, is this annual or monthly ?" <<<

I am not knowledgable about real estate issues, but did discuss the cost issue with the president of a small company that I used to work for.  Going over the numbers can be useful:

Cubicle: 8' x 8' = 64sq.ft. (large, but your example)

Private office: 10' x 12' = 120 sq ft. (Not 4x the cubicle, but quite adequate in my experience).

Lease rate: $12/sq.ft. / year.

Multiplier: 2.  You have to figure in some factor to account for hallways, bathrooms, etc.

Cubicle/year: $1536
Office/year:  $2880

Difference: $1344
Difference as % of $80K/year salary: 1.7%

The above is based on actual numbers for one of my employers.  But I am not an company accountant, so there might be more to add.  Anyone have better numbers?

Wednesday, May 1, 2002

one issue with the cost of private offices, is that one has to find an office building that has a bunch of private offices. if you cant find that, you have to build out offices yourself, which I'm guessing is very expensive. just leasing or buying the cube farm equipment is relatively expensive.

the best scenario I have seen, is everyone work at home, then come in on thursdays for meetings...

ghetto superstar
Wednesday, May 1, 2002

Windows is the world's most sophisticated PC platform and is in use on more than 90 percent of desktop computers. Microsoft is one of the world's most successful and richest companies.

The engineers who develop Microsoft's products each have their own office. Enough said.

Hugh Wells
Wednesday, May 1, 2002

Mathematically, work accomplished is something less than proportional to people thrown at a project.  2 people may only get 1.5x as much done as 1 person.

The natural argument here is that money spent increasing productivity buys you more than spending that money on another person.  So, the 2% increase figure for an office is really less than that, considering the alternative.

If you want software quality over several software versions, you'll have a difficult time getting it if you have a mercenary army that turns over every 18 months.  If you're doing an in-house system, it's less critical.

Give people a good environment and keep them around.  Worth 2%?  It's worth 20 times that.

Bill Carlson
Wednesday, May 1, 2002

Joel tells us: "Tom said that groups of people who work in the same style on the same kinds of tasks will not distract each other."

I hate to say anything disparaging about Tom after he wrote such a wonderful book, but I think this will only be true after human cloning has been perfected.

Counter examples from my experience:

In a shared two person office my office mate, working on the same project, likes to listen to some really obnoxious radio talk program.  Nothing as intelligent as Rush Limbaugh.  Just a lot of yammering, dinging of bells, and yelling in irritating squeaky false voices.  Yes, that's distracting.

In a cubicle/quadricle farm it's easy to overhear conversations one or two cubicles over.  Personal conversations, sometimes about family matters and other things unrelated to the project, start up occasionally (too often) and are a distraction.

In a three person office it seemed like at least one of us was always on the phone or meeting with someone else who had come in to the office.

In some sense I think Tom's statement is right, but as a practical matter I don't think any group of more than one person will always be working in lockstep to the point that they don't cause each other significant distraction.

Wednesday, May 1, 2002

I take this issue very seriously partly because I read Peopleware, believed it, got a 22 office space for my workers, and watched how they were about 1/3 as productive as other places I've worked with everybody in an open space.

My current belief is this is an issue of the type of programming you are doing.  If you are doing some kind of programming with fairly well thoughtout designs and goals then you can take your spec/document back to your office and code away with no interruptions.  In that kind of programming I'm guessing offices work better than cubes or open space.  But, I work in the gaming industry and the thing is, it seems like the most important thing for making a great game quickly is team interaction.  That doesn't mean having a meeting and deciding Artist A, make a dog, Designer B, put it in a level, Programmer C, make it bark and growl.  It means all those people in the same room looking over each other shoulders, encouraging each other, giving each other constant feedback.
When I look around the room and I see a new part of the game that doesn't look very good I immediately want to comment on it as in (it could be better if it was this way) or (what about if you did this).  That kind of feedback is invaluable for games where the goal is not just meet the spec but also make something visually exciting and fun.  Two things you can't spec out or at least I've never seen them spec'ed out.

I'm not trying to say just gaming is different but it's a matter of style.  Sure I can get distracted in an open room.  When that happens I generally put my headphones on as does everybody else.  On the otherhand having had the office situation and seen the barrier it puts up to quick team communication I'm now totally 100% against office for the type of work I do.  Maybe the guys that build the libraries could have offices but the team members need to be together.

That's my two cents.

Gregg Tavares
Thursday, May 2, 2002

Nat Ersoz, what were you trying to do with your post apart from advertise thomsons? you got some products shipped without having private offices. so what? i could just as easily say "they would have been done much more quickly if everyone had had an office free from from distraction". your "argument" is empty.

Thursday, May 2, 2002

Ahh boy,  dare I answer a troll?

>> Nat Ersoz, what were you trying to do with your post apart from advertise thomsons?

Wrong.  I now work for a start up, which I'm sure you've never heard of, and by all indications is going to compete with Thomson head on.

>> i could just as easily say "they would have been done much more quickly if everyone had had an office free from from distraction".

No you couldn't. First, and most obviously, you were not there.  Second, these were first ever products.  First ever cable modem, first ever MPEG deployment.  Something which I used as examples for how an open "cube" environment enabled development.

You, however, offer no examples of how offices enable development.  On the other hand, I offered a stinging example of a product that never shipped: MSTV.  An example of how offices, I hypothesised, helped crush that work.


Nat Ersoz
Thursday, May 2, 2002

I'm thinking that hardware-related development is not as abstract as development of complex software, and that there is perhaps not the same value in intense concentration for prolonged periods?

Studies have been done on productivity, and they show unequivocal improvements for offices, for intellectual tasks such as software development. I'm afraid I don't have references to hand.

WebTV flopped for reasons unrelated to offices.

Hugh Wells
Thursday, May 2, 2002

Reading the anti-office postings has been interesting.  It is hard to believe that anyone other than a bean counter would think that shared office space would be more productive than private offices.  Of course, I know that some types of development work are quite different from what I do, but it is surprising that there is such a great difference that some types of work would be less productive.

As a software developer I am concerned about productivity, but as an individual I am also concerned about something a little more personal.  I want a job where I can enjoy coming to work in the morning.  I know people who can tell you how many years, months, and days they have until retirement.  I don't want to live like that.

At one time I did have a job where I enjoyed coming to work.  Private offices were only one factor.  It was well managed, it was an interesting project, we got a good product out and after about 15 years it is still in use.

On another project we were put in dimly lit cubicles, the project did get done in about twice the time originally scheduled, but when I walked in from the parking lot I would look at the building each morning and dread going in.

Thursday, May 2, 2002

Would people in offices say that mailing lists are vital?  One of my little fantasies is to transition our workplace to private offices + openspace, but one clear change is that we must use mailing lists, or other suitable 1-to-many communication.  Might some of these failures have been caused by not changing communication at a deep level?

Thursday, May 2, 2002

">> i could just as easily say "they would have been done much more quickly if everyone had had an office free from from distraction".

No you couldn't"

i already did. you quoted it, remember? now prove that it would not have been true.

you say that those projects would not have been delivered in the same timeframe if everyone had had private offices. you have not produced one iota of proof to support that. you say that project 'x' (in this case mstv) failed because people had offices, but again offer no proof.

your attempt to extrapolate "offices are bad" from "i have successfully developed products in cube space" is flawed. maybe they were successful because the engineers and management were good. maybe because the coffee was free (generates goodwill from the employees), or maybe because it was not free (ppl don't waste so much time yattering around the coffee machine).

i bet you that i can reverse your own "arguments" to "prove" that offices mean better productivity based on projects _i_ have worked on. (clue: some others of us have worked on projects in both private and open offices and might have different tales to tell. can i really be productive in an office of 30 or more where mobile phones are constantly ringing?) i would say that Daniel Shchyokin has it spot on - "it depends". it depends on management, culture, and whole slew of other things.

Thursday, May 2, 2002

>> I'm thinking that hardware-related development is not as abstract as development of complex software, and that there is perhaps not the same value in intense concentration for prolonged periods?

Baloney.  Obvioulsy one whose never written VHDL or Verilog before.  Hardware is much more difficult than software (though I'm a software engineer, not hardware).

>> WebTV flopped for reasons unrelated to offices.

Do tell...

Perhaps it flopped because DirectShow is a poor architecture for multimedia development.  Or that WinCE performance is so abismal.  Or that MSFT delayed deployment in order to grab server market share and alienated their customers.  Or that MSFT doesn't know how to work in a joint deployment environment.  Or that MSFT shipped crap to their hardware manufacturers which was DOA.  Take your pick I suppose, there are lots of variables.

If MSFT is the model for offices (i.e. MSFT = good :: office == good), I'd say I've proved my point.

Nat Ersoz
Thursday, May 2, 2002

Nat, so you don't want to do dinner with Bill?

Always good to hear points of view. The truth lies in there somewhere.

Hugh Wells
Thursday, May 2, 2002

In reading through these posts, I get the impression that someone is just searching for another silver bullet and believe they have found it in a private office.  Although there are probably some development tasks that could benefit from non-interruption, those tasks would probably be better accomplished away from the main workplace, and that doesn't have to be a private office.

IMO... most of the people I have worked with over the years yearn for the prestige of the private office as status symbol, not to improve their own productivity.  In fact, most spend little to no time watching how they work and care even less how to improve. 

Yes, there are bound to be exceptions, but if the majority of your co-workers have their productivity as their number one concern over the feel good of what other people think of them, then you truly work at an exceptional shop.

Joe AA.
Thursday, May 2, 2002

Some replies to Joe AA.

>>>"someone is just searching for another silver bullet and believe they have found it in a private office."<<<

I agree with Brooks that there is no silver bullet.  But that doesn't mean there are not better or worse ways to do software development.  We are just arguing that the benefits exceed the costs.

>>>"probably some development tasks that could benefit from non-interruption, "<<<

You do different work from what I do.  I'd say just about all.

>>>"most of the people I have worked with over the years yearn for the prestige of the private office as status symbol,"<<<

That may be true for you.  In the best environment I ever worked in everyone, except the receptionist, had a private office: developers, administrative people, accountants, everyone.  Who would the prestige be relative to?  Maybe to developers who worked at lesser companyies?

>>>"if the majority of your co-workers have their productivity as their number one concern over the feel good of what other people think of them, then you truly work at an exceptional shop. "<<<

Most people I have worked with are concerned about their career, but are generally dedicated to producing quality software "on-time and under budget" if possible.  They don't measure productivity by lines of code, but by getting the program running right.

Thursday, May 2, 2002

In this on-going discussion of private offices vs shared office space, we need to keep in mind that an important factor is noise.

It is easy to describe the ideal software development office as four walls, a door, a window and basic office furniture. But if there is excessive noise, the benefits can easily be wiped out.

On one project I was working as a contractor using a shared office space.  It was somewhat noisy and the desk would vibrate from the HVAC system.  One day I wanted a quiet space to work for a couple of hours, so I went in to one of the private offices normally used by employees, but was temporarily vacant.  The HVAC system in that office made so much noise it felt like I was sharing it with a tornado.  After about five minutes I gave up and went back to my vibrating desk.

When we try to argue that private offices are superior development environments, we really need to say "quiet private offices".

Thursday, May 2, 2002

If I can turn off my phone when I'm in the flow, and if no one will ever walk in to my cube and start talking as though I could just stop what I'm doing and give them my full bandwidth and immediately swtich back when they leave, and if everyone else in the room had their conversations, phone or otherwise,  in a way/place that I can't hear them, and if the clue-free hypersocials who don't understand that a programmer can be at their most productive while staring at the wall or walking around the block or otherwise doing something besides frantically typing will be kept out of the room (as much for their benefit as mine), then I don't need an office.

Dan Sickles
Thursday, May 2, 2002

swtich = very fast switch

Dan Sickles
Thursday, May 2, 2002

If you don't have an office space with separate offices there is no need to despair.
There are cubicles and cubicles.

Our UK office recently moved from a space with mostly private offices to entirely cubicles. However they spent a lot of money
on good cubicle dividers - eight feet high, well soundproofed, good quality furniture.
Also separate areas (divided by solid walls) into which 4 or 6 people would go. Those of us who liked private offices were sceptical at first, but it actually works well.

There are also two other factors in the cost of cubicles versus offices. With cubicles you need more meeting rooms per person. Also, if someone offered me a choice between a pay rise of $1500 and a nice working environment, I would take the nice working environment any day. Thata's not enough money to be miserable for eight hours a day.

David Clayworth
Thursday, May 2, 2002

I've found that I get used to background nois, its the constant flow of people into my cube, has anyone ever tried, putting up a do not disturb sign and then angrily yelling, complaining to ones manager when they are disturbed

Daniel Shchyokin
Friday, May 3, 2002

Headphones are a good symbol. Years ago when I worked in open plan, I would sometimes put the headphones on without music, just to tell people I was busy. They called it being in deep download.

Again, many years ago, I did actually reprimand senior managers once. They had come into a development area for a meeting and were standing around chatting at high volume, self-importantly, breaking everyone's concentration.

I went and asked them to quieten down. They did.

Hugh Wells
Tuesday, May 7, 2002

Has anyone considered the lower costs of lighting and air conditioning in a office environment?  Also, what about the lower occurance of passing colds around in a office environment?

Steve Woodson
Friday, January 16, 2004

*  Recent Topics

*  Fog Creek Home