5 ms·

Bootstrapped, Profitable, & Proud: GitHub

238 points by fogus 16y ago

joelmichael 16y ago
Rock stars living the dream.
delano 16y ago
This is an interesting observation: The first year or so of the company reminded me a lot of the awkward teenage phase of self-discovery. GitHub the company had sort of sprung up from this side project, so we never had any big vision or dream or aspirations. We just wanted to work on something cool. I’d love to say that’s all you need, but we’ve learned there’s more: you need to have a vision and a philosophy. Everyone (all the founders, at least) need to be on the same page. The hard part is finding that page. Do we make web apps, or just do source control? What do we pay our employees? Should we speak at conferences? How do we approach customer support? I don't hear many people talking about the value of a company vision and philosophy. I occasionally hear people talk about company culture, but usually in an abstract way.
- DTrejo 16y ago
  They never really said how they manage this, or whether they wrote it down. What ways of keeping people on the same page are most effective?
  - cschep 16y ago
    Great question, also it would be fun to hear about what "page" they ended up on. They have an incredible service that I love using, so what did they decide to actively NOT do?
  - bobf 16y ago
    If you haven't already, take a look into Zappos and Tony Hsieh for great insights into conveying company culture.
    - brown9-2 16y ago
      To elaborate: in "Delivering Happiness" Tony talks about how Zappos goes as far as publishing a real Zappos Culture book each year, containing stories from hundreds of employees/partners/vendors about what Zappos the company means to them. The book is a great read - amazing how well they've institutionalized the idea of a company culture and methos for holding onto it as they grow.
  - subwindow 16y ago
    I have a feeling it was mostly accidental. Having a bunch of awesome people increases the likelihood of it happening, but I don't think that is a very repeatable lesson.
    - werrett 16y ago
      Having a bunch of awesome people increases the likelihood of it happening... I'd hazard a guess and say the non-repeatability comes down to the awesome bit. It is easy to label someone as awesome after coming out on the good side of a rough patch. It is much harder to be comfortable in that label when sizing up a potential co-founder.
      - derefr 16y ago
        If you assume the set-equivalence that "awesome people get on the same page quickly" implies that "people who get on the same page quickly are awesome", you then find, by http://en.wikipedia.org/wiki/Aumann%27s_agreement_theorem http://en.wikipedia.org/wiki/Aumann%27s_agreement_theorem, that "awesome" means rational. The question is whether the assumption holds :)
al1 16y ago
fogus, thanks for highlighting that article, which then showed up in the HN feed on Twitter. Great advice and perspective from Chris Wanstrath.
jacquesm 16y ago
That was worth reading just for this bit: "We generally ignore the advice/opinions of others as a rule. GitHub is the company of the compelling argument – every decision needs to stand on its own merit. That someone successful (or unsuccessful) tried something before might matter in a discussion, but what matters more is how the idea itself applies to the situation."
- goof 16y ago
  Every company would describe itself as a meritocracy though. In github's case they're probably still small enough for that to be true.
  - jacquesm 16y ago
    Agreed, but in this case I actually believe them. As soon as you've got two layers of management that probably is the first to go out the door.
    - derefr 16y ago
      I've never understood why a startup would choose to grow to the size where it needed two layers of management, unless the individual product they were working on was really of a complexity too high for a small team to manage (like, say, a nuclear reactor.) For products that stay the same size, but gain new sub-products, cross-products, side-businesses, etc., why not just "undergo mitosis" and become two startup-sized companies, each handling a single facet of the business but sharing a treasury and building?
      - jacquesm 16y ago
        There was a company (BSO) that did just that, they split every time they got to a certain number of employees (I believe the magic number was 50). They were hugely successful. You can read all about it here: http://www.extent.nl/articles/entry/the-evolution-of-bso-and-its-offspring/ http://www.extent.nl/articles/entry/the-evolution-of-bso-and...
jmcnevin 16y ago
I want that octocat plush doll.
- nuclear_eclipse 16y ago
  I would pay for one.
- riffraff 16y ago
  I thought it went by the name octopussy, but i want one too
  - graywh 16y ago
    http://en.wikipedia.org/wiki/Octopussy_(character) http://en.wikipedia.org/wiki/Octopussy_(character)
- technomancy 16y ago
  ITYM pentacat.
bhellman1 16y ago
It's people with this type of personality that change the world.
codexon 16y ago
One thing Tom had learned at his previous venture, Gravatar, was that offering a resource intensive service for free was a losing proposition. Is anyone else wary about the continued existence of gravatar? Even though it was bought out by Automattic, I think it might shut down in the future because it is resource intensive and hard to monetize.
- gojomo 16y ago
  I would think it could be optimized to be quite efficient, and breakeven value might be possible via the data it throws off -- every Gravatar is a web bug.
n8agrin 16y ago
The most salient point to me was the fact that github was not initially conceived of as a business, but as a tool the founders desired for the betterment of their own day to day. And not only that, but github was quickly adopted in the founders' workplaces at the time, giving them time to focus on it and refine it without giving up an income. Dogfooding is a common theme in startups and it's worth noting that 37signals' Basecamp product was, as I understand it, built with the a very similar purpose. Initially it was a tool for them to help manage their own projects first and something they could charge for second, built under similar conditions where they already had a business designing client websites, providing them an income. A lot of times I dream up projects I think of as potential startups but often have to ask myself, "Would I actually use this product?" If the answer is ever even a hint of "no" I tend to suspect that I could not dedicate myself to the project wholeheartedly enough to back a business. I wonder if this is a meaningful metric and if dogfooding is really such a powerful motivator.
- GrandMasterBirt 16y ago
  Is that not how most great technologies are born? Ok some are for war, but most of the "cool" technologies I've seen (even some made by co-workers) are all about someone having a major itch and developing the best damn back scratcher they could to scratch that itch. Its quite incredible what someone with a vision can do, even without technical prowess that some posess. Its all about just thinking that "I am only one man, I only have a few hrs a day... however I need to do the job of 10" and instead of giving up building something that lets you do just that. I find the in corporations where you can throw more people/time at the problem is where it starts to fail. I can always "hack" something up fast, and add a few programmers to make it faster, but the trick is to not add anyone and still meet the deadline (or a bit over, but that gets covered by blowing all other deadlines away). GitHub is no different. They had an itch, scratched it, turns out other people had the same itch.
muhfuhkuh 16y ago
Wow, the cottage industry that is Linus Torvalds keeps churning out the goods that make empires (Red Hat, Android) and up-and-comers (github). "I'm not a businessman; I'm a business, man!" --Jay-Z
davidw 16y ago
Paradoxically, I think some of these businesses 37signals is listing are more interesting than they are themselves. Part of their 'unfair advantage' is simply being famous. Github is famous too, because they took a smart approach to the free/paid split that works out pretty well.
lzw 16y ago
All to often, people seem to associate "not taking venture capital" with being a "lifestyle business". GitHub is an interesting business. Like EngineYard or any other tool oriented startup, its potential is not the same as a broad consumer internet business (like Facebook or Google). But I don't think Github will be a "lifestyle" business. It already isn't. I don't think Paul Graham has a problem with this. But it seems like a lot of his followers do / might. I think there's an important distinction here that is missed and I'm not sure how to bring it up. (Last week I asked about doing a startup outside the US, because the US is not an option for US, and one person presumed we were doing a "lifestyle business" because we expressed costs as one of our concerns.)

238 points by fogus 16y ago

Rock stars living the dream.

This is an interesting observation: The first year or so of the company reminded me a lot of the awkward teenage phase of self-discovery. GitHub the company had sort of sprung up from this side project, so we never had any big vision or dream or aspirations. We just wanted to work on something cool. I’d love to say that’s all you need, but we’ve learned there’s more: you need to have a vision and a philosophy. Everyone (all the founders, at least) need to be on the same page. The hard part is finding that page. Do we make web apps, or just do source control? What do we pay our employees? Should we speak at conferences? How do we approach customer support? I don't hear many people talking about the value of a company vision and philosophy. I occasionally hear people talk about company culture, but usually in an abstract way.

DTrejo 16y ago
They never really said how they manage this, or whether they wrote it down. What ways of keeping people on the same page are most effective?
- cschep 16y ago
  Great question, also it would be fun to hear about what "page" they ended up on. They have an incredible service that I love using, so what did they decide to actively NOT do?
- bobf 16y ago
  If you haven't already, take a look into Zappos and Tony Hsieh for great insights into conveying company culture.
  - brown9-2 16y ago
    To elaborate: in "Delivering Happiness" Tony talks about how Zappos goes as far as publishing a real Zappos Culture book each year, containing stories from hundreds of employees/partners/vendors about what Zappos the company means to them. The book is a great read - amazing how well they've institutionalized the idea of a company culture and methos for holding onto it as they grow.
- subwindow 16y ago
  I have a feeling it was mostly accidental. Having a bunch of awesome people increases the likelihood of it happening, but I don't think that is a very repeatable lesson.
  - werrett 16y ago
    Having a bunch of awesome people increases the likelihood of it happening... I'd hazard a guess and say the non-repeatability comes down to the awesome bit. It is easy to label someone as awesome after coming out on the good side of a rough patch. It is much harder to be comfortable in that label when sizing up a potential co-founder.
    - derefr 16y ago
      If you assume the set-equivalence that "awesome people get on the same page quickly" implies that "people who get on the same page quickly are awesome", you then find, by http://en.wikipedia.org/wiki/Aumann%27s_agreement_theorem http://en.wikipedia.org/wiki/Aumann%27s_agreement_theorem, that "awesome" means rational. The question is whether the assumption holds :)

al1 16y ago

fogus, thanks for highlighting that article, which then showed up in the HN feed on Twitter. Great advice and perspective from Chris Wanstrath.

jacquesm 16y ago

That was worth reading just for this bit: "We generally ignore the advice/opinions of others as a rule. GitHub is the company of the compelling argument – every decision needs to stand on its own merit. That someone successful (or unsuccessful) tried something before might matter in a discussion, but what matters more is how the idea itself applies to the situation."

goof 16y ago
Every company would describe itself as a meritocracy though. In github's case they're probably still small enough for that to be true.
- jacquesm 16y ago
  Agreed, but in this case I actually believe them. As soon as you've got two layers of management that probably is the first to go out the door.
  - derefr 16y ago
    I've never understood why a startup would choose to grow to the size where it needed two layers of management, unless the individual product they were working on was really of a complexity too high for a small team to manage (like, say, a nuclear reactor.) For products that stay the same size, but gain new sub-products, cross-products, side-businesses, etc., why not just "undergo mitosis" and become two startup-sized companies, each handling a single facet of the business but sharing a treasury and building?
    - jacquesm 16y ago
      There was a company (BSO) that did just that, they split every time they got to a certain number of employees (I believe the magic number was 50). They were hugely successful. You can read all about it here: http://www.extent.nl/articles/entry/the-evolution-of-bso-and-its-offspring/ http://www.extent.nl/articles/entry/the-evolution-of-bso-and...

jmcnevin 16y ago

I want that octocat plush doll.

nuclear_eclipse 16y ago
I would pay for one.
riffraff 16y ago
I thought it went by the name octopussy, but i want one too
- graywh 16y ago
  http://en.wikipedia.org/wiki/Octopussy_(character) http://en.wikipedia.org/wiki/Octopussy_(character)
technomancy 16y ago
ITYM pentacat.

bhellman1 16y ago

It's people with this type of personality that change the world.

codexon 16y ago

One thing Tom had learned at his previous venture, Gravatar, was that offering a resource intensive service for free was a losing proposition. Is anyone else wary about the continued existence of gravatar? Even though it was bought out by Automattic, I think it might shut down in the future because it is resource intensive and hard to monetize.

gojomo 16y ago
I would think it could be optimized to be quite efficient, and breakeven value might be possible via the data it throws off -- every Gravatar is a web bug.

n8agrin 16y ago

The most salient point to me was the fact that github was not initially conceived of as a business, but as a tool the founders desired for the betterment of their own day to day. And not only that, but github was quickly adopted in the founders' workplaces at the time, giving them time to focus on it and refine it without giving up an income. Dogfooding is a common theme in startups and it's worth noting that 37signals' Basecamp product was, as I understand it, built with the a very similar purpose. Initially it was a tool for them to help manage their own projects first and something they could charge for second, built under similar conditions where they already had a business designing client websites, providing them an income. A lot of times I dream up projects I think of as potential startups but often have to ask myself, "Would I actually use this product?" If the answer is ever even a hint of "no" I tend to suspect that I could not dedicate myself to the project wholeheartedly enough to back a business. I wonder if this is a meaningful metric and if dogfooding is really such a powerful motivator.

GrandMasterBirt 16y ago
Is that not how most great technologies are born? Ok some are for war, but most of the "cool" technologies I've seen (even some made by co-workers) are all about someone having a major itch and developing the best damn back scratcher they could to scratch that itch. Its quite incredible what someone with a vision can do, even without technical prowess that some posess. Its all about just thinking that "I am only one man, I only have a few hrs a day... however I need to do the job of 10" and instead of giving up building something that lets you do just that. I find the in corporations where you can throw more people/time at the problem is where it starts to fail. I can always "hack" something up fast, and add a few programmers to make it faster, but the trick is to not add anyone and still meet the deadline (or a bit over, but that gets covered by blowing all other deadlines away). GitHub is no different. They had an itch, scratched it, turns out other people had the same itch.

muhfuhkuh 16y ago

Wow, the cottage industry that is Linus Torvalds keeps churning out the goods that make empires (Red Hat, Android) and up-and-comers (github). "I'm not a businessman; I'm a business, man!" --Jay-Z

davidw 16y ago

Paradoxically, I think some of these businesses 37signals is listing are more interesting than they are themselves. Part of their 'unfair advantage' is simply being famous. Github is famous too, because they took a smart approach to the free/paid split that works out pretty well.

lzw 16y ago

All to often, people seem to associate "not taking venture capital" with being a "lifestyle business". GitHub is an interesting business. Like EngineYard or any other tool oriented startup, its potential is not the same as a broad consumer internet business (like Facebook or Google). But I don't think Github will be a "lifestyle" business. It already isn't. I don't think Paul Graham has a problem with this. But it seems like a lot of his followers do / might. I think there's an important distinction here that is missed and I'm not sure how to bring it up. (Last week I asked about doing a startup outside the US, because the US is not an option for US, and one person presumed we were doing a "lifestyle business" because we expressed costs as one of our concerns.)