As someone who is remote about 90%, I contend the watercooler chat magic is overrated and a waste of time. I was previously in office full time. I was one of the first tech hires. I am mostly a programmer, with a good deal of overlapping ops and analysis thrown in. I'm part of a very small engineering staff (under 10) in an overall small company (under 30 people). My work output in our open office drops to around 10% of the output when I work from my house. This is code, documentation, and meetings attended. People spend more time at the office because they are constantly interrupted (both by off-topic conversations as well as for work-related things) and need to stay late (past 40 hours) to finish their tasks. From informal notes, conversations that occur via Skype or In Person are poorly documented and tend to be rehashed more often than written communications via Github or Slack. IMO, if the entire team was remote and utilized more written communications, we'd be spend less time rehashing old topics and more time delivering value. As to the spontaneity and ideation of face to face, I've observed this happening over video chat and slack just as often as in the office. If people are excited about their work, they will talk about it, in person or not.

I agree 100%, this matches my experiences pretty closely. I worked 80% remote for a six months a while back, and it was astonishing how much better productivity was, as well as the quality, if not necessarily the quantity of communication. Now I've basically flipped that ratio, and it is depressing how much of my time is wasted by low-quality communication, interruptions, as well as the low-grade stress and irritation of having to be in the office.

> At least in the startup world, most people are concerned that you will miss that zeal and watercooler chat magic that is critical to early stage companies.

Is there any real evidence showing how much this actually matters, or is it just Valley voodoo? I suppose that is partially tied up in the answer to your question.

> I don't think anybody cares too much about your 1000th engineer being remote, the organizational DNA has already been set a long time ago.

Yet hundreds of companies at this size seem to care shitloads about it.

I worked at a node.js startup Strongloop before it was purchased by IBM (https://www-03.ibm.com/press/us/en/pressrelease/47577.wss)

We were primarily a remote work company as part of the DNA - it was great and productive.

We were small and mostly senior technical folks so the level of trust was pretty high.

People were located in the Czech republic, the Netherlands, Vancouver, Minnesota, China, and Cali.

Best work experience I have had by far.

Scrapinghub is 100% remote from day zero. Nowadays there are ~140 people spread around the world, covering almost all timezones.

I can't find a reference anywhere for this to back it up, but I recall reading that Stephen Wolfram (https://en.wikipedia.org/wiki/Stephen_Wolfram) started Wolfram Research (Mathematica) with 100% remote employees.

37Signals (now Basecamp)

Were they really remote in the beginning? Remember, they started as a "New Media" agency in the late 90s, and built up their following in their blog before pivoting into SaaS. It's hard to imagine having a creative agency all remote, and everything I've read from them—and I have been reading them since that early Zeldman era where they first rose to prominence—indicates that they started hiring remote to find the best people to work on their core products, not for client work.

Not that I don't think you can start a good startup as all-remote, but I just don't think that's how 37S bootstrapped.

Zapier if I recall correctly.

Stack Exchange (now Stack Overflow) has a lot of remotes. I'm not sure what the exact ratio is, but it's not too far off from 50/50 (in engineering, anyway).

Automattic (creator of wordpress) is fully remote, I believe.