Yes, you've come across the core criticism of the VC business model. They fund 500 startups in the hopes that one of them goes big. Not big, gigantic! Humonguous! The next Google! Facebook! Dropbox! They need such a huge return for their business model to work (do you know how expensive it is to fund 500 pie-in-the-sky startups?)
With that in mind, if as a founder you're happy to settle for mid-size, you're going to get into headlong conflict with VCs who've funded you (and thus own majority shares of the company), because you're failing them according to their model.
You may be interested in jwz's and idlewords's comments, who frequently decry this model.
It's not about going public, it's about raising VC. So don't raise VC. But then the question is: how will you fund product development before you become profitable? If that's not a problem for you, you're in the minority.
You can raise VC without giving up majority control.
In two ways.
First, don't allow dilution to push you + any co-founders below majority. If you focus on building a profitable business, and you're not trying to be the next Uber, this is extremely doable. Ellison at Oracle, Gates + Allen at Microsoft, and Omidyar + Skoll at eBay were able to basically do this; Bezos & Family were also roughly able to accomplish this leading up to Amazon going public (his family still owns nearly 20%).
Second, use the Page-Brin voting shares control approach. Most investors will not like this, don't take their money. Find investors that sync with your vision for the company.
I took on a well known investor, and gave up a modest corner of my business. I retain majority, and will continue to indefinitely. The key is to be able to grow your business into the second or third inning with that first investment, then strictly control dilution thereafter. For example I could do two rounds of 15% dilution from here, and the investors would still not have majority control.
If your business needs to bleed large amounts of red ink, then you're going to put yourself into a corner accordingly. There probably is no good outcome unless you happen to be among the tiny number of companies that make it out of that start alive. For everyone else, you always have a choice.
I've always thought the VC model for software is fundamentally flawed.
Software is so easy to get started that if it can't bootstrap from a bare minimum product, maybe the idea isn't that great as a business.
On top of that, I think there's a bit of circular logic and wishful thinking involved in getting VC money. It encourages companies to spend more money than they need to on things like scaling and fancy design and excessive marketing, because they need to scale to meet VC expectations.
The VC model was essential for the hardware oriented companies that pioneered it because they needed a lot of equipment and capital to build chips and computers. There's so little capital cost with software nowadays, I don't know why VC still persists.
That is definitely not true of all software. Consumer and SMB Web applications, maybe, but domain-specific business software can be a rather nontrivial and capital-intensive undertaking. Those are the quiet companies engaged in the "boring" business of making money that you don't hear about.
That said, a lot of the money goes to marketing. Google AdAwords and low-touch web-based marketing doesn't work for all kinds of products; in an esoteric business market niche, the cost of customer acquisition is rather high, as trade shows, conferences and personal connections matter a lot more.
I don't think it's as hard as many people are making it out to be. Not every company is going through huge growth phases super early on, and that's where the need for huge piles of VC cash right now on any terms comes from. If you keep your costs low by growing carefully and maintaining at least some revenue while also being very selective and careful about the investments you take I think a lot of companies can do just fine even without being 100% purely bootstrapped.
There is an entire eco-system of startups outside the little typically SV/VC bubble where everyone is in it to disrupt everything and win big, there are tons of small companies growing at reasonable levels that will never be the next google or facebook. There's nothing wrong with a "mere" multi-million dollar company.
He had an incredible amount of early traction. For every startup/VC dyad, there's going to be a power imbalance. Either the founder(s) and their company are in high demand (rare), or the VC is (typical).
The side that has the leverage is going to get the best terms. Normally that's the VC. (Liquidation preference, your first born, board seats, etc.)
Should founders just take less investment if their initial product isn't wildly popular? Tho burning VC money to do market exploration (and failing) doesn't sound too bad (isn't that what it's for? :)
I'd love to see more nascent startups follow the GitHub path, and seek product-market fit before attracting investment.
Aaron puts on his speculation hat
GitHub didn't need VC funding. They were incredibly profitable from selling their enterprise product, and took funding because one or more of the following were true:
1. The founders wanted liquidity. Like, buying a Gulfstream or three liquidity.
2. The founders were getting bored and wanted to put the 'fuck you, we IPO'd' notch in their collective belt.
3. Early employees with equity were getting restless about their lack of liquidity and threatening to quit unless they got theirs.
I've got another hypothesis for GitHub: the dynamics of the market changed in a way that it became winner-take-all, which then made it make sense to take VC.
When GitHub started, they were a developer tools company, and the common wisdom was that developer tools have very little lock-in, there's room for multiple players in the space, etc. Sometime at just over a million users, that assumption ceased to be true; you now get companies that host on GitHub because all their prospective employees are familiar with it, and users that sign up for GitHub because their employers use it. With that network effect, it makes sense to grow as quickly as possible to exploit the network effect, and if they didn't take VC, they'd risk some other startup doing that.
It's not just GitHub: pretty much everyone in the developer tools market started taking VC right around then. StackOverflow in 2010; Atlassian in 2011; GitHub in 2012. In Atlassian's case, they were founded in 2002 and didn't take VC until much later. Also, developer-oriented startups started getting funded, eg. Parse, Meteor, Codecademy & other bootcamps.
I understand the antagonism towards VCs, but they have financial reasons why they push for only unicorns. Best reasons I have heard so far are:
1) The only way VCs can make large amount of money for their investors is via a liquidity event. Investors would much rather prefer .25% of a 1B dollar public company than 25% 10M dollar private company, because you can liquidate the former far more easily.
2) Overhead costs are significantly less when you focus on fewer companies, which is what happens when you push all companies to be a unicorn. A VC would much rather care about 5 companies with significant promise than 500 small companies. In the latter situation, you would need way more employees to make sure your investments are safe.
3) Funding a few unicorns is better for marketing purposes. VCs need to raise money from institutional investors, hence need to "wow" them.
With that in mind, if as a founder you're happy to settle for mid-size, you're going to get into headlong conflict with VCs who've funded you (and thus own majority shares of the company), because you're failing them according to their model.
You may be interested in jwz's and idlewords's comments, who frequently decry this model.