[cletus]

Working from home is a popular topic for us engineers. Generally speaking there are two broad camps in this argument:

1. Those who want to work from home and argue largely from that position. They say they'll be just as productive, you can write code from anywhere, having a more flexible schedule will make them happier and more productive and so on; and

2. Those that think there is more to your job than the lines of code you write. In even small companies (maybe even especially small companies?) culture is important. Culture transmits largely by physical proximity. Osmosis if you will. There is value in team camaraderie, whiteboard sessions, going to lunch with colleagues, sometimes just sitting around and shooting the breeze about whatever.

I fall very firmly into camp (2). This also applies to splitting teams geographically (common within Google) and, all other things being equal, you're better off having your organization in N locations versus N+1 locations.

Raises and promotions are more a function of relationships than anything else. Not being there decreases visibility and diminishes relationships. Or perhaps it's just that those who basically just want to write code see no value in and/or spend no time on building relationships?

So I think if you found a group of likeminded people that just wanted to put their heads down and write code then they could probably work together as an effective distributed team but as soon as you're in the minority in that situation you're losing out and (IMHO) it's not what's best for a colocated team anyway.

[adrianhoward]

Generally speaking there are two broad camps in this argument

There's at least one more camp. Those who think that there's large body of evidence that, all other things being equal, a good co-located team will be much more productive than a distributed one.

And I say this as somebody who lives in a relatively rural part of the UK, runs their own company, and spends a pretty large chunk of my time telecommuting and working with other remote workers (because, often, all other things are not equal :-)

For example:

http://conway.isri.cmu.edu/~jdh/VRC-2008

"It doesn't take much distance before a team feels the negative effects of distribution - the effectiveness of collaboration degrades rapidly with physical distance. People located closer in a building are more likely to collaborate (Kraut, Egido & Galegher 1990). Even at short distances, 3 feet vs. 20 feet, there is an effect (Sensenig & Reed 1972). A distance of 100 feet may be no better than several miles (Allen 1977). A field study of radically collocated software development teams,[...], showed significantly higher productivity and satisfaction than industry benchmarks and past projects within the firm (Teasley et al., 2002). Another field study compared interruptions in paired, radically-collocated and traditional, cube-dwelling software development teams, and found that in the former interruptions were greater in number but shorter in duration and more on-task (Chong and Siino 2006). Close proximity improves productivity in all cases."

http://www.springerlink.com/content/0137yud7c3k8xryw/?MUD=MP

"Based on the empirical evidence, we have constructed a model of how remote communication and knowledge management, cultural diversity and time differences negatively impact requirements gathering, negotiations and specifications. Findings reveal that aspects such as a lack of a common understanding of requirements, together with a reduced awareness of a working local context, a trust level and an ability to share work artefacts significantly challenge the effective collaboration of remote stakeholders in negotiating a set of requirements that satisfies geographically distributed customers"

http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?reload=true&#...

"Our results show that, compared to same-site work, cross-site work takes much longer and requires more people for work of equal size and complexity. We also report a strong relationship between delay in cross-site work and the degree to which remote colleagues are perceived to help out when workloads are heavy"

http://kraut.hciresearch.org/sites/kraut.hciresearch.org/fil...

"Our findings reveal that: software developers have different types of coordination needs; coordination across sites is more challenging than within a site; team knowledge helps members coordinate, but more so when they are separated by geographic distance; and the effect of different types of team knowledge on coordination effectiveness differs between co-located and geographically dispersed collaborators."

http://www.computer.org/portal/web/csdl/doi?doc=doi/10.1109/...

"One key finding is that distributed work items appear to take about two and one-half times as long to complete as similar items where all the work is colocated"

http://possibility.com/Misc/p339-teasley.pdf

"Our study of six teams that experienced radical collocation showed that in this setting they produced remarkable productivity improvements. 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.Teams in these warrooms showed a doubling of productivity"

I could go on....

There are certainly arguments for telecommuting being more productive if you have a bad onsite working environment. My ideal solution is to fix the bad environment if at all possible, rather than distribute the team.

[mindcrime]

I could go on....

Yes, I wouldn't argue against the notion that - in general - co-location is better for (at least short-term) productivity. The question, to me, is "what about the other effects?" Does forcing co-location decrease job satisfaction, for example, and what's the effect on turnover? Or what about morale, and it's impact on productivity over a longer period of time?

There are also issues like "what if the super talented developer I want to hire will only work remotely?" Am I better off forgoing his/her talents completely, or accepting a (perhaps less than optimal) remote work arrangement?

I don't think any of these issues have black and white answers; and I think that determining the ideal work environment is still a bit (art|black magic|luck|etc) and not yet a science.

[adrianhoward]

Agreed. As I said "often, all other things are not equal" :)

That said - it's generally been my experience that people are happy when they're being most productive, and vice versa. The most productive, gelled and happy teams I've worked with have all been co-located (which does not mean I think distributed teams are unhappy or unproductive).

I think many folk see teleworking as a solution to a bad onsite working environment. Often those people are not in a position to "fix" their normal working environment and it's the only option they have to be happier and more productive. More power to them.

Unfortunately, since poor onsite working environments are common, when those people are in a position to build their own teams, working environments and businesses they often miss the opportunity to build a great productive onsite environment - and then miss out on the productivity (and general happyness IMHO) wins that can get you.

[marknutter]

"Those that think there is more to your job than the lines of code you write. In even small companies (maybe even especially small companies?) culture is important. Culture transmits largely by physical proximity. Osmosis if you will. There is value in team camaraderie, whiteboard sessions, going to lunch with colleagues, sometimes just sitting around and shooting the breeze about whatever."

You're waiving your hands here; flailing, even. None of this is quantifiable, it's all your based on your gut feeling. There are plenty of ways culture can transmit online, but I'm afraid my examples would involve just as much conjecture as yours do.

I telecommute, but also make a trip to be on location every couple of months. I agree that face to face contact is important to some degree, but not in any way I could accurately measure. I'm FAR more productive when I'm telecommuting because I don't have colleagues interrupting me with questions ever 10 minutes, I don't feel the temptation to "shoot the shit", I can work through lunch, and I waste no time commuting. As a result I'm more productive than the average non-telecommuting worker.

Raises and promotions may be a function of relationships, but perhaps they shouldn't be? Perhaps they should be based on performance.

[davidw]

> You're waiving your hands here; flailing, even. None of this is quantifiable

Just because it is not easy to quantify does not mean there isn't some truth to it. 'Flailing' and 'hand waving' seem exaggerated.

[stfu]

Oh if only social interations were that easy quantifiable. Life would be a lot easier. We could avoid all wars, be full of love and full of love and happiness, developing only the most successful products without ever failing to understand the (irr)ationality of our clients. Social data is so subjective, as it is almost impossible to control externalities, and something like the (un)importance of "facetime" is in my opinion always circumstantial. Therefore I would agree with the gut feeling comment, as it comes really down to the individual employee to evaluate his intentions/priorities/relationships and what works best for that situation.

[adrianhoward]

As a result I'm more productive than the average non-telecommuting worker.

Although sometimes optimising somebody's individual productivity gets in the way of optimising the productivity of the company as a whole.

You need to figure out where the bottlenecks are before you know the right place to optimise :-)

[mindcrime]

Although sometimes optimising somebody's individual productivity gets in the way of optimising the productivity of the company as a whole.

That's the real rub. It's entirely possible that the Right Thing To Do - from the perspective of the firm - is the do something other than what is most optimal for the individual. Sp, as a lot of us geeks tend to be fairly individualistic, and because it can be hard to isolate the variables when you're talking about team productivity, a lot of us fall on the side of "optimize for the individual."

As somebody who's radically individualistic, and who is also a startup founder, I find myself torn on the idea of whether or not it even makes sense to have an office, or whether it makes sense to push for a completely distributed team.

I'm getting a bit of a real-world experiment with these issues now, as my $DAYJOB has me in Chicago for a 6 month consulting gig and my co-founders are back in the RTP area. We use email and IM heavily now... but we had an in-person "hack day" last weekend when I was home, and there was definitely value in all 3 of us being in the same room, huddled up together.

I'm leaning towards thinking firms probably should have offices, and that most people should be in the office at least part of the time. But I would take a pretty laissez-faire approach towards it, I think. Provide offices, but give people fairly unlimited freedom to work remotely or come into the office as they see fit.

[adrianhoward]

I'm leaning towards thinking firms probably should have offices, and that most people should be in the office at least part of the time. But I would take a pretty laissez-faire approach towards it, I think. Provide offices, but give people fairly unlimited freedom to work remotely or come into the office as they see fit.

That sounds like a good idea.

I'd maybe consider running some experiments too. See what happens when you have everybody work in the office for a month. See what happens if you have everybody working remotely for a month. I've had clients try things like that and be surprised by the results.

[theorique]

Raises and promotions may be a function of relationships, but perhaps they shouldn't be? Perhaps they should be based on performance.

Not saying you're wrong - performance is a great metric to use for gauging contribution to an organization. However, I find that proposals that begin with "people should" or "groups should" are often about as easy to implement as cat herding.

[Angostura]

As someone else pointed out just because something is difficult to measure doesn't make it unimportant.

One part of your post intrigues me.

> don't have colleagues interrupting me with questions ever 10 minutes

Let's assume that your colleagues aren't idiots. That they need these questions answered. Either they are having to go away and ask someone else, or they are having to figure the answer out themselves (potentially taking a lot more time than if you had simply given the answer) or they are deciding to make do without the answer.

Each of those is potentially detrimental to the organisation's effectivenesses as a whole. Your "productiveness" is more than simply the sum of the code that you produce.

Moreover you're missing all those 'interesting little problems' that turn up, that you would know the answer to, if anyone only thought to ask you.

You're at risk of diminishing your total value.

(And yes, I primarily teleworked for several years).

[marknutter]

That makes sense on the surface, and there's certainly value to spreading knowledge, but when you're in the same room there's a much lower barrier, so the tendency is to ask questions first, rather than making a serious attempt at solving the problem on your own. Make no mistake, we still ask each other questions when we're in a bind, but this communication is asynchronous. We ping the other person over chat, but that person is free to respond in their own time. I can't count the number of times I've gotten an IM from a colleague asking a question, only to get a "nevermind, I figured it out :)" message 5 minutes later. To me, that's a much more productive workflow for both individuals and the organization as a whole.

[xutopia]

It was clear which camp you fell into just by the way you defined the two.

[_ea1k]

This post is hilarious when read a bit more accurately. Basically, you have said that there are two broad categories, rational people and irrational people. I guess it's a nice technique for argument, but I suspect that most people do not fit nicely into either of your buckets.

I want to work from home, and do enjoy doing so. I also like working from the office and enjoy doing so. There's more to my job than lines of code that I write. Culture is important. Culture sometimes transmits via physical proximity, and sometimes other means (IM, Phone, etc). Sometimes one or two meetings a week (or even month) is more than adequate to cover the parts that don't communicate so easily remotely. Sometimes remote workers go to lunch (or dinner) with colleagues. Sometimes my whiteboard is a computer.

Your group 2 covers a broad range of people and needs, some of which may sometimes require (or significantly benefit from) close proximity, and some of which often won't.

Basically, people, personalities, and working style vary.

[SCdF]

We have a tester who lives in another city from us, and I've managed to build camaraderie with him using nothing but skype IM, including breeze shooting and other non-work based activities. I have done the same in previous jobs, with people on the other side of the world. I don't think you need to be in the same room.

[seany]

I was just going to say more or less the same thing. I think I have a better working relationship with one of my coworkers that lives 1500mi from me than the guy I sit next to. Now I'm sure that personality plays a part here, but it's clear that distance really isn't `that` much of a detractor.

[countessa]

In large companies that I've worked for, the uptake of IM and skype is slow and barring the occasional video conference when someone just couldn't travel for one reason or another, is almost frowned upon.

That kind of makes it difficult for remote working to take place effectively, because, whatever else we may think, work is more than just "give me my list of tasks and I'll give you the deliverables".....the social interaction and relationships between people can radically improve motivation and such.

[TimGebhardt]

It's funny that in the companies where working from home I think is more likely (big companies with policies about it and such), it would probably be less effective because those same companies probably have locked down computers and firewalls and the users wouldn't be as able to use what they might deem most effective (Google Docs? Skype? Dropbox?)

Whereas with smaller companies where you can pretty much do carte blanche ("install whatever software you need on your PC as long as it doesn't drain our Comcast business internet"), working from home would probably be looked down upon.

[w1ntermute]

As is often the case, the best solution is a compromise. A policy of allowing 2 or 3 days/week of working from home would allow for "real" work to get done from home (not to mention the gas and time savings) while also having the advantages of physical proximity.

[adrianhoward]

As is often the case, the best solution is a compromise

Nope - there isn't a best solution.

It depends on the company, the project, the team, etc. Sometimes everybody telecommuting is best. Sometimes a mixed model is best. Sometimes an everybody-onsite model is best. Sometimes flying everybody to a common location for a month is best. Sometimes...

[prof_hobart]

>and, all other things being equal, you're better off having your organization in N locations versus N+1 locations.

All other things aren't equal, though. There can be less interuptions at home, there's less time wasted on commuting, and more flexibility - allowing (for instance) people with children to have a better work/life balance, which should hopefully make them more effective workers.

There's also things like the simple cost of the building - if 20% of your staff are typically at home on any given day, you need 20% less floorspace etc (I know it's not quite that simplisticaly linear, but you get the point).

For meetings, things like Facetime can be remarkably effective at getting someone "in the room" when they are physically at home. It's not quite as good as them actually being there, but it's remarkably close.

Obviously there's a balance. There are quite definitely upsides to having people in the building - a quick 2 minute chat at someone's desk with a piece of paper and a pencil can sometimes be as effective as a Visio diagram and an email that took an hour to write, and even more importantly having people sat right next to you can make working far more collaborative.

But depending on what your job is, having the flexibility to work for a day or more a week at home can be hugely valuable.

[gaius]

Reminds me of a conversation at my last job:

Manager: Working from home is a privilege, not a right Me: Having me work from home is a privilege not a right

You see it cuts both ways. If management wants more than the 9-5 with an hour for lunch that it stipulates in your contract, that is.

The truth is the technology for working remotely just isn't there yet. Sure you can just churn out simple code that way, might even be faster than in the office. But nothing beats the sheer mental bandwidth of a dozen engineers in a room with big whiteboards and an ample supply of coffee. Not even IRC!

[gcp]

Culture transmits largely by physical proximity. Osmosis if you will.

Culture transmits by communication, and nothing else. If you don't have a good (or equal) communication with your remotes, things are going to suck.

If you have a good communication with your remotes, they'll absorb culture faster than the most introvert people that are in the actual office.

Your post essentially lacks any base for the claims you make.

[mbesto]

I think this is better served by saying that culture transmits by communication, in all it's forms, which include body language, physicality, etc.

[SkyMarshal]

>culture is important

I see this asserted a lot, but what exactly is meant by 'culture', and is it really true?

For example, if by 'culture' you mean having an office full of well-educated, capable, motivated engineers and problem-solvers vs not well-educated, capable, or motivated, then sure, I buy that.

Strong leadership that intimately knows the business? Sure, Intel and Apple being exhibits A & B.

But beyond that, what parts of culture should every business strive to replicate, and what parts are more likely overvalued in the zeitgeist due to survivorship bias and other cognitive mistakes?

[hack_edu]

Have you ever worked somewhere that was perhaps a decent job, but you dreaded coming in to work because of bosses/coworkers/distractions? That's culture, just bad culture. Everyone seems to be able to relate to bad culture. Company culture is a two way street, good culture seems harder to come by than bad. I can't even say I've even been somewhere with 'good' culture but it definitely exists and is really just the inverse of bad culture.

I'd say one of the things that draws people to telecommute most is too much exposure to bad company culture. Arguments in regards to productivity and environment are more responses to 'bad' culture than anything else.

[Spooky23]

Think of culture as cooking.

You can use the finest meats, freshest vegetables and highest quality herbs, and still produce an awful meal by assembling them wrong. But on the flip side, you CAN make a very good meal from mediocre ingredients by preparing them properly.

Two examples form my career:

Workplace A: full of smart, motivated problem solvers and "A players" who failed in key ways because a toxic culture just poisoned anything. People were nasty and busy protecting fiefs. The individual work efforts were great, but they couldn't come together.

Workplace B: A place where there was a fairly broad range of people from marginally skilled folks to the super-smart where the teams just got along and had leadership that promoted cross-training and cooperation. People were friendly & professional.

As a leader, doing the care & feeding to maintain good culture (aka making work a place where you want to work) is hard. I think that after you reach a certain size, 100% remote workers make it harder, at least for a core team. Also, your business requirements around security and other factors may force you to do things that will damage that remote culture.

[r0s]

It's not such a hard dichotomy. I think working from home occasionally makes good sense, if you treat it like what it is, a productive break.

Just the fact that so many request it, should show the benefit of releasing that pressure occasionally. I hate working with tired, overworked people on mental cruise control. The fact is, a big part of the workday is filled with social interaction, whiteboard sessions, etc. But all that is just as draining (if not more so for some) compared to the actual work tasks.

I'd also like to point out the huge time waste that is commuting.

[goostavos]

>I'd also like to point out the huge time waste that is commuting.

It's also massively stress reducing. No daily traffic? It's awesome. I've been working from home for an out of state company for about 3 years now. I could make more money if I switched to a different job, but not having to commute has actually become a massive perk for me.

Having little stress, being able to relax at home, and still be very productive are worth more than monetary compensation to me.

I've got a family member that makes about 15k more than I do, but spends 3 hours commuting each day! That more of less eats up all of her leisure time through out the week. Then weekends are house chores. There is nothing about her job that couldn't be done from home. 3 hours in traffic each day is no way to live..

With no commute, I wake up, do some stretching and light exercise, drink some coffee, and then when I'm ready to begin, I just sit down at my desk and get started.

Worth every penny that I don't make.

[loceng]

The rule I am getting comfortable with for people I'm hiring: I don't care where or when you get work done, just get it done. If that means home, if that means office, if that means a cafe - go for it. If that means at 3am - awesome, be productive when you're in the state of mind, mood and place to be productive. I'm primarily a creative person, as most good engineers are - problem solvers, and this is how I end up being most productive.

[einhverfr]

Things cut both ways though.

On one hand transmission of ideas etc. in real time is important. On the other, face time means lots of additional distractions, so many employees may be more productive working from home.

Culture is important too but I don't see why that can't be done over remote aspects as well. Certainly the Linux project has a different culture than the PostgreSQL project, and these have different cultures than the LedgerSMB project, and in all of these you are dealing with geographically dispersed teams.

So the question is, how do people interact? How can you help them interact? How can you limit their interactions to productive ones? I think that remote and face-to-face environments just pose different challenges in these areas.

[lsc]

>Raises and promotions are more a function of relationships than anything else. Not being there decreases visibility and diminishes relationships.

Eh, while I agree that this is how it /is/ - this is certainly not how it should be. (I mean, management is a function of relationships, sure, but if you base Engineer pay on relationships, which many companies do, you are going to end up with a bunch of worthless sycophants.)

[laymil]

Based solely on my own experience, I believe that the benefits to be gained by physical proximity decrease according to an inverse square law. The benefits gained from sitting with the team you're currently working with can't be discounted, but as teams grow and your team overflows a single office, the synergies shared between team members in the same room far outweigh those with team members in the next room, to the point where they could be next door or across the country with little difference.

Embracing the distributed approach and realizing the cutoff point leads to building better tools for collaboration and integrating them into the work environment. Without this recognition, the mentality is that you can always talk to someone in person.

I've worked in multinational companies where collaborating with someone across continents, bar the time difference, was just as easy as working with someone down the hall. I currently work for a company that has not yet achieved this recognition, so working with anyone outside my immediate office takes hard work and determination.

[dagw]

While I agree with what you say and also put myself in camp (2), the one advantage about splitting teams geographically is that it puts you geographically closer to more customers. Depending on the nature of your business it can be very handy to meet customers, both current and potential, face to face from time to time, and having people in the same cities as your main customers makes that a lot easier.

[Spooky23]

You also have a camp of people who wish to practice their craft, period.

If your goal is to move up the organization, being a remote worker is similar to working in the Iowa field office -- you're not making the personal connections with the bigshots that you need to make to get the promotions. Being in the right place at the right time and luck are big success factors, imo.