How can they have done this economically whilst bootstrapping?
I find it hard to see how companies can make money when they have to pay for data traffic and storage.
A repository isn't likely to cross the 1 gb threshold easily. They also email you if it does.
According to A16Z's podcast or something somewhere they have been profitable since inception so it was not much of a problem.
Keep in mind traffic and storage are cheap, and cheaper still in bulker quantities. 1 TB of data is about $30 a month on S3. If each repository averaged 50 mb one user's $7 a month subscription could support nearly 5,000 repos. 50mb is also quite large for a repository btw.
If each was about 5mb that's nearly 50,000 repositories that could be supported cost wise with their cheapest $7 micro plan.
https://enterprise.github.com/home
https://www.crunchbase.com/organization/github/funding-round...
After that, they grew enough to keep up with growth in the company and features until they got the Andreseen investment and threw growth into overdrive.
I can almost guarantee that I am using less than 5Mb throughout my Github, so I don't doubt they are making money.