Bootstrapped GitHub Now Raising a Round from Andreessen Horowitz
pandodaily.com
pandodaily.com
These questions about GitHub vs. Pinterest are revealing in that I think many folk on HN don't realize what a relatively small community of developers this is. Relative to the tens of millions of folk on Pinerest. Heck, GitHub's front page reveals they have only 1.6 million users.
I wonder if their growth curve has started to flatten out. I hope they have a great plan/vision for what to do with the money, because running GitHub ad infinitum as an immensely successful small software company would seem to be a dream to me.
Nevertheless, I'm inclined to agree about the apparent popularity of git vs its actual popularity. I don't think I've ever met anybody who's used git professionally, and in fact I don't even know that many people who've actually used git at all. Everybody I meet (though I haven't met absolutely everybody!) uses perforce if they can afford it, and SVN if they can't. SVN supports binary files tolerably well, and perforce supports them fine. perforce keeps working fine even when your depot's head revision (if that's the right term) gets to 300GB. Well... SVN probably doesn't do that. But it's still cheap.
But people in games probably wouldn't use git anyway on account of its poor handling of binary files, and limited scalability in the face of gigabytes of stuff. (I have a suspicion it might be a bit difficult to explain to many/most artists anyway. And a fair proportion of programmers too...)
Of course, this shows nothing more than that we both live in our respective little bubbles ;) But my impression, as a bit of an outsider, is that the HN hivemind is made up of the same people that frequent GitHub. That is, mainly (but of course, not exclusively) people who work with web-python-ruby-javascript-etc. - thus naturally magnifying the apparent importance of GitHub. But I suspect the HN population is not representative, and therefore GitHub's popularity amongst the HN readership probably does not extend to the programming population at large. And this is certainly backed up by my experience.
(All standard disclaimers apply.)
- How is GitHub worth less than Pintrest?
- What could GitHub possibly gain from raising money? It's clearly not about money. Is this a liquidity event? Couldn't they just issue a dividend for that, they've clearly got plenty of cash… Or is this about the credibility of having a valuation. In that case, couldn't they have gone public? Amazon went public at a sub-$500M valuation, and look at them now!
Edit: come to think of it, I can see the possibility of wanting to acquire other companies, and not currently having the capital to do so. That's about it.
I'd imagine VC would help them create much better support options for the enterprise (the company I'm at usually don't even consider products without commercial support), such as 24/7 local support, support contracts etc.
[1] enterprise@github.com
My suspicion is the founders are taking money off the table in a similar manner as 37signals did when they took money from Bezos Expeditions [Jeff Bezos]. Considering the fact that they're bootstrapped, profitable and growing like a weed it seems reasonable to sell a portion of the company to the greatest advisor [Marc Andreessen].
Going into enterprise means more than having just a behind-the-firewall solution. Deploying code behind the firewall means that you are at the mercy of the customer's environment and have to be much more forgiving than running in an "opinionated" SaaS environment.
To wit, here is a laundry list of things you have to deal with for enterprise customers, and they all require time, money, and support.
+ Slower release cycle than SaaS - you may not be able to push updates to customer environments at will. They may require that they decide when to take updates, which means their codebase could be several months behind. (Not always, but it can happen).
+ Data & privacy laws, particularly for European customers. Do you love S3? A lot of companies don't, and governments especially don't. This means potentially rolling your own elastic file storage which must also be deployed behind the firewall.
+ You get into managing a cluster for the customer, not just a single instance.
+ Your software must fit the security scheme of the customer. In a SaaS environment, you typically have a great deal of control over the network topology. Behind the firewall, company policy may require that only certain ports be open. This might mean re-writing code to either serve a lot of protocols over port 80 or being on long support calls while people file tickets and wait for IT to punch open holes for you. (And then sometimes close them right back up, requiring you to repeat the cycle every time you deploy).
+ It's also not uncommon to have a dedicated team just for the enterprise customers, or even a dedicated team just for one enterprise customer. This requires hiring people who must be trained and brought up to speed. As I'm sure you're aware, engineer salaries are a big cost factor.
+ If time differences are too large or too inconvenient, you may need to hire people in other parts of the world. This means needing to comply with local laws and regulations. Which usually means hiring lawyers, additional HR people, etc.
The list goes on and on. Deploying for the enterprise can be much more involved than what you might think.
And I guess you think Newspaper profits (used to) come from the price people pay for them.
The cost of ads has plummeted in the last few years, and I seriously doubt Pinterest would remain profitable if its primary revenue stream is ads.
Google's search advertising has shown targeted advertising is one of the most profitable business models ever. In some ways Pinterest is even more targeted than search.
He raised the money from Benchmark because he needed help recruiting executives to help run and expand the company. His recruitment efforts before that had been largely unsuccessful, and he openly admitted that no one took him seriously before he raised money.
Yes, the internet was a new and unproven phenomenon at the time, but there are other reasons to hitch up with a VC other than money.
The other factor is hype. Pinterest is the next hot social company and their comps are Facebook, Twitter, LinkedIn, and Instagram. Easy to let your imagination go wild if you're an investor.
re: #2 - they would raise money to a) grow faster, b) acquire companies, and/or c) provide shareholder liquidity. b & c are self explanatory, so let's focus on a):
The wad of cash lets them hire faster. Let's say they have 70 employees today and they want to grow to 200 by year end (for enterprise support, for example). Hard to triple your salary expense that quickly if you're operating off cash flows. GitHub have a network-effect business, so why not lock-in every company on the planet as quickly as possible? It'll lead to a much, much greater profits in the long-run.
It's only worth the dilution if you expect your equity to be worth more than 1/(1 - n) [See PG's essay: http://paulgraham.com/equity.html]
hype investment is a poor investment, unless you are looking to cheat the late comers to the 'hype', instead of looking to create value (in which case, the investor is a blood sucker, and should be hated).
Poor investments looses society wealth. In other words, it should not be that a hyped up fad of the month company garners more investment monies than a real, profitable company.
I'll speculate.
Github's fighting a losing game. They are trying to charge for something they are also giving away for free, and their target market is people who are uniquely qualified to replace them. There are plenty of quality Github alternatives around right now. To stay on top they have to continue to be at the forefront development-wise and they have to continue to have the majority of developer attention to reap the network effects. It's precarious.
Compare to Pinterest. They make money, ostensibly, from every product someone pins. People pin lots of products, and the people doing the pinning and the browsing have higher conversion rates than people browsing Google or other sites looking for things to spend money on. Pinterest is also a brand; it will be hard to get everyone using Pinterest to switch just because a better version exists somewhere else. People can collaborate with Git without using Github; the only tangible loss is the pull request feature.
It takes Pinterest less work to convert a byte of bandwidth into profit. Not only do they need less development effort moving forward, they also stand better odds of making money on some random user.
So something is surely off.
"It takes Pinterest less work to convert a byte of bandwidth into profit. Not only do they need less development effort moving forward, they also stand better odds of making money on some random user."
But they also need many more users to actually turn it into revenue.
Contrast this with early stage companies (which is what we're talking about here). VCs don't care as much if a company has a proven business model - what they're looking for is GROWTH opportunity. The opportunity to completely disrupt a huge market or invent a new one. While profitable companies have better odds of surviving long enough to get to that point, they may not be the best investments in the end.
To give you a concrete example of how this could affect investor decision making. Let's say that the total opportunity for Source Code Management solutions is $2bn/year in corporate spending (completely making this number up) - even if GitHub captures 100% of this market, which they won't, they're capped at $2bn/year in revenue, unless they figure out another way to make money. Once they're done paying the sales guys, engineers, and otherwise supporting operations, they make maybe $500mm/year in profit. With a standard P/E multiple, they'd be worth something like $5-$10bn.
Pinterest, on the other hand, has the potential to be the "facebook of online shopping" to amazon's "google of online shopping". If people really engage with pinterest, and they spend a lot of time on the site, and they're more likely to buy stuff when they're looking at what their friends like and are buying, then it could be worth a lot more than that. The point is, while they don't yet have predictable revenues or earnings, their potential market (online shopping) is so much bigger than github's, that it's possible to justify a high valuation based on growth potential.
I'm not saying I buy the Pinterest hype. I'm just saying that there are criteria beyond current profitability and cash flow on which one can value an early stage business.
The differences between Pinterest and GitHub are that: GitHub is more unique (it's not strictly social media and there's more homework to do); it's bootstrapped and negociates from a stronger position; VCs want to give start-ups more money than they need because that's how they get paid.
Github however is profitable and has a proven, functional business model. Github is absolutely not fighting a losing game, because they're providing a service companies obviously desperately want to outsource (thus the mass usage). Providing that service in a secure manner is worth a lot of money, as Github is protecting something extraordinarily valuable.
There are plenty.
There are a LOT of people using Pinterest, and a LOT of Amazon products on pinterest.
1) The assumption is that the affiliate link model and market will scale to the size of Pinterest's eventual traffic maximums, that's a big assumption.
2) The assumption is that Amazon will want to allow Pinterest to suck profit out the door at that scale, instead of Amazon competing with Pinterest in one form or another instead, just like they're now competing with Groupon / Living Social (despite an investment in LS).
3) Amazon has a very long history of directly competing with companies that touch their platform.
That's an extraordinarily shaky business model to say the least.
I'm sure if I can't think of ways that Twitter and Tumblr can make outsized profits from their massive respective user bases, then I'm not being creative either. Oh snap, some of the best business minds are already on the problem and they can't figure it out either.
The same holds true for Pinterest, they're stuck in the same monetization box, and they're not going to suddenly show massive profitability either. It's blatantly obvious that platform will never monetize to the scale of their traffic.
Having said that, switching costs are low because of the distributed nature of Git. Not sure what would happen if something genuinely better came along.
If something genuinely better comes along, the network effects would still be a huge source of inertia given the FUD around switching.
switching costs are low because of the distributed
nature of Git
Lock-in is not measured only in regards to the potential switching costs. Lock-in is also possible by providing the best possible service. Along with real network effects (people recommending your service to other people) then the global lock-in becomes inescapable, even if the theoretical switching costs are low.
And quite the contrary, GitHub locks people in simply because most developers have a GitHub account. People use GitHub these days to put their portfolio on display. And I've worked with a couple of companies that manage their projects on GitHub and it was a pretty painless process to send me an invitation for joining their private repos.
Even Linux is mirrored on GitHub and even Linus Tolvards has a GitHub account. GitHub is not only a really good Git hosting service, but in case you haven't noticed, it's also a vertical social-network built for developers. And their service keeps improving by leaps and bounds, compared to their competitors, which are merely followers. I'm so happy with GitHub that even though I can have as many private projects as I want on Bitbucket, I still pay GitHub for the privilege to have private projects.
Yes, something better can always come along. But just as Google did for search engines, with GitHub around it's pretty hard to build something better. And this to me is the best possible kind of lock-in and the sign of a thriving and growing company: the product being popular, not because it's forcing users, but because it's simply the best.
Would you rather have stock today in MySpace or a hypothetical 5-year-old SubversionHub? Obviously, the answer is different for different people, and there are good reasons to support and use Github and good reasons to invest in it. That programmers are good at product loyalty simply isn't one of them.
Github is a decent medium-term bet, but Pinterest is a much better lottery ticket.
Would you rather have stock today in MySpace or
a hypothetical 5-year-old SubversionHub
This is the criteria after which I judge companies:
(+) is the product solving a real problem for me? (yes)
(+) are they a sustainable, profit-generating business yet? (yes)
(+) is the product forcing users to remain customers? (no)
(+) are they smart and hard-working enough to adapt to new market conditions? (yes)
(+) do they still have growth potential? (yes, because GitHub is lately a social network for professional developers, being preferred over LinkedIn by many)
That programmers are good at product loyalty simply
isn't one of them.
The same goes for GitHub. Even if I stopped being a GitHub customer, I would still go to GitHub because most popular projects are managed on GitHub or have a mirror there. Popularity in turn helps GitHub get valuable feedback, which it can use to further improve my experience, which in turn makes me a happier customer and so on.
The fact that the top two comments are saying this is a sure sign we are progressing into a full-fledged bubble.
See George Zachary's interview: http://techcrunch.com/2012/05/17/in-the-studio-crvs-george-z... -- it sure seems like we are entering "Stage 2" (check out the video, not the article)
When we start discussing what we feel one company is worth based entirely off of the valuation of another, unrelated, presumably inflated valuation, it's the beginning of the end.
Ditto for Groupon, Zynga, Pandora and the like having the insane valuations that they did, while generating zero profit - their valuations have imploded massively. Or a company like HomeAway, that was sporting a $4 billion market cap, and a 600 pe ratio; or LinkedIn with its 1,000 or so pe ratio. The bubble has already exploded for companies like Netflix, whose valuation has collapsed back to earth. A few more like CRM are temporarily defying gravity.
On the other hand, if it's all VCs and funds of funds losing silly-money, then no harm no foul, I suppose.
For their last quarter (NY Times): "Zynga said its net loss was $85.4 million, or 12 cents a share, in the most recent three months."
Groupon has lost a lot of money the last four quarters. Their latest quarter could best be described as break-even, and negative with option expenses.
Reuters: "Groupon reported first-quarter pro-forma net income, which excludes option expenses, of 2 cents per share, versus a net loss of 41 cents a share, a year earlier."
No doubt the right direction. However, a company that's break even, with $2.x billion in annual sales, is not worth $8 billion. It's a lot better than the $20x billion they were formerly worth of course.
a) Media companies are generally pretty valuable. Pinterest allows users to show clear interests, which should be very easy to convert to commercial intent.
b) The potential market for Pinterest users is much larger, with much less competition than GitHub
c) Software development tools generally are difficult to make money on. Infrastructure can be slightly better, but still expensive. A16Z knows this market as well as anyone (http://www.hp.com/hpinfo/newsroom/press/2007/070723xa.html)
Facebook's share price is very high compared to their current revenue. There are very valid questions about how quickly they can grow revenue, and those questions are reflected in the share price.
Markets aren't rational.
Here is a link to the talk, and a relevant quote from the introduction:
http://www.justin.tv/startupschool/b/272031754
"But for me, I don't have to worry about these things, because GitHub has never taken any funding, ever. So I want to talk a little bit about how you can avoid this mess of VC, if you so choose, by telling you a little bit of my story."
2- They already avoided most of the problems of VCs by getting to profit.
"The ironic thing about bootstrapping and venture capital is that once you demonstrate some success, investors will come to YOU. When this happens you will be in a much better place to make a more reasoned choice about taking on additional capital and all the complexities that come with it. Talking to VCs with some leverage in your back pocket is an entirely different game from throwing yourself in front of a conference table full of general partners and trying to persuade them that you're worth their time and money. Power is happiness."
[1]: http://tom.preston-werner.com/2010/10/18/optimize-for-happin...
I think the bootstrapped B2C game is ALL about saving money/not going broke. It's most accurately about making great decisions (it always is, at any level), but I think for what they were doing, that meant not spending. Their management/founders might have been great at that. I think it's very much an engineer's mentality, anyway.
However, B2B is a different animal, I think, and one that Tom Preston-Werner might not be geared to hunt -- at least not until someone turned him on to how much bigger GitHub could be. It's not cheap to get the attention of big companies, neither in pursuing them at all, nor in buying the people you need to pursue them -- engineers usually can't do this by themselves. You need veterans of _that_ game and it's a different game (at least from what I've seen).
This makes the Instagram deal look even more out of proportion. Github ought to be worth a lot more, it's a profitable huge business with real assets, used by millions everyday for work.
But the question is: what could they do with more money that they can't do now?
Proper enterprise support requires customer representatives, phone lines, sales people. All that is expensive and requires capital.
Of course, this is just a wild guess. I have no firsthand information.
But we still pay for it, because what else is there? :)
That might be wrong, but I'd need to see a lot of evidence first.
[1] http://blogs.atlassian.com/2011/11/why-we-removed-wiki-marku...
(Which now supports Git: http://blog.bitbucket.org/2011/10/03/bitbucket-now-rocks-git...)
There's also the option of Gerrit which puts most of this in a single WAR (not gitweb, though as with gitolite the integration is documented), with a code review workflow on top.
Disclaimer: I work at Fog Creek.
I worked for a smaller outfit that sold to enterprise sometimes directly and sometimes through larger partners (CA primarily). Except when we told them they had no idea of our size.
Some specific gripes included very slow support response time and poor options for installs (vm based only). There were others but it boiled down to the ops guys asked to set up and maintain the system weren't happy with the way they were treated.
I don't think 'going enterprise' is good for the culture of any company, but if they want to do it, they definitely have to kick up their game.
Trying to win on purely technical merits won't work. Being unresponsive especially during an evaluation is the kiss of death.
Git and github-style content-management requires a mental model very different from the mass/mass-enterprise market... but the tech-leaders way of thinking will reach there over time via repeated exposure, and Github is doing a great job of providing the benefits in approachable steps, and a whole package of related modern hosting/distribution/authoring assistance.
Why wouldn't presentations, spreadsheets, and all other forms of writing eventually be amenable to dvcs-style revision control?
What would the repercussions be from taking cash from Andreessen Horowitz? Will the original Githubber's keep their control over the company? I'm honestly a little worried, but at the same time if they want the cash to expand and make Github better, then I am all for it.
Where have they said they were against it? I think they are often held up as an exemplar of bootstrapping by other folks, but I don't recall too much from them directly either way.
My list might be a bit frivolous, but money would give options, at least.
(Edit: Actually, why not Heroku style cloud hosting too? The code's already on the service ready to roll..)
* github managed to hire & acquihire top talent like crazy
* their UI/UX people really focused on making github's UI more consistent and usable
* they spent time to go native
* the enterprise solution got a lot of love
so they have huge momentum and spend their time very well. personally I think they should look into making jobs.github.com THE resource to hire great talent. In a way github already is very relevant in the hiring process but similar to carreers.stackoverflow there is huge untapped opportunity to disrupt the hiring market.
Anyway I think github has a bright future ahead and I'm very curious about what will happen in the coming months.
It's not correct.
As many has said it, Github > Instagram + Pinterest. I would say at least 5B$.
In addition, if founders want to take cash off the table and that is what it takes to make a great investment in github you do it.
A VC wants 10x, I see no better way to allow this than to let a founder take some money off the table and shoot for the moon.
As far as risk is concerned with expansion, as long as the founders and the VC want the same outcome I dont see what the big deal is with taking on more risk, particularly for founders with other peoples money. Even better after theyve taken money off the table!