Hacker News Reader

Comment by moritzplassnig | Hacker News Reader

When GitHub's financials got leaked in 2016, GitHub's personal accounts accounted for 12.5% of their revenue. GitHub Enterprise was 50%, their organization (= business) accounts were the remaining 37.5% [1]. It's fair to say that the personal accounts are irrelevant from a revenue perspective.

[1] https://medium.com/@moritzplassnig/github-is-doing-much-bett... (disclaimer: I wrote the article but the data is by Bloomberg)

View on HN

> It's fair to say that the personal accounts are irrelevant from a revenue perspective.

Your assertion can only make sense if you also believe that a 12% paycut is irrelevant to your income.

Revenue != income. It's quite likely the profit margin on developer accounts is lower than the other types, and if this cuts the knees out from under the competition may well be cheaper than, say, paying for advertising.

But that is a non-sensical argument unless you are saying that making all those accounts free will somehow reduce the overhead related to the personal accounts. Seems like it will in fact increase costs as more people sign up.

>if this cuts the knees out from under the competition

All of Github's competitors (i.e. gitlab and bitbucket) already offer unlimited private repos. This is Github catching up.

Exactly,while it is dramatic for github, it has to do this because of competitions.

I was thinking that Bitbucket still has its promise to always be free.

However, I don't think it will be an issue. As hosting costs drop to zero, if someone starts charging, they lose customers to a free competitor. (as is happening here).

As well as MS’s own Azure Devops aka VSTS as well as free hosted builder servers, private package repos, project management, etc. for up to five users.

12% of Github's income is ~0.001% of Microsoft's income.

12% of Github's income is 12% of Github's income.

12% of GitHub's income is 100% of 12% of GitHub's income.

$20 is $20

true === true

If I’m reading the revenue chart correctly, 12.5 percent of their revenue was approximately 11 million dollars in 2016. At Microsoft, who is even going to notice that...

There’s surely a goodwill factor and wider strategic benefits here beyond the relatively small amount of revenue - these individuals enjoying the free private repos will in many cases likely be future candidates for customers for the enterprise offerings etc, or one day find themselves involved in discussions regarding managed VCS for a larger org. Given Microsoft’s current strategy to win over developers, this seems like a pretty sensible move, and the price (for Microsoft) very low. Strategically this seems not entirely different to the free accounts AWS throw at university students. Once you have someone hooked on your free offerings, you can start locking them into the wider platform (Azure) much more easily. Again, pretty much the AWS playbook.

Also, let’s not forget the 12.5 percent figure comes from 2016, a year which saw some pretty significant changes to GitHub’s offerings, such as unlimited personal repositories for paying users among other things. Prior to this you had to pay more to get more private repos. The number today may be somewhat different - I know my github spend went down following that change.

Also, 12.5 percent of revenues - comparisons to a 12 percent pay cut are pretty misleading.

Comparing 12.5% revenues to a pay cut is not misleading at all.

Unless there is some reason to believe that making individual accounts free will somehow also reduce costs in a material way. Seems to me it will likely increase costs significantly as many more users will flock to create free private repos.

In fact, it's more like giving you a 12% percent pay cut and then asking you to commute further to work.

Focusing solely on the impact to Github’s bottom line kinda misses the entire point of Microsoft wanting to acquire them in the first place.

A decision like this has to be viewed in the wider context of Microsoft’s market strategy and the rest of their developer tool portfolio, of which Github is just one piece.

If Microsoft just cared about short term gains, there were much better ways to blow 8 billion dollars.

But saying 12% is irrelevant to REVENUE is focusing on revenue. You can say losing the 12% is worth it, but you can’t say it’s irrelevant to revenue.

Indeed -- seems to me that the recent MS moves with Github are part of the same thought process that led them to include Bash / WSL in Windows 10.

> Unless there is some reason to believe that making individual accounts free will somehow also reduce costs in a material way

It's misleading because it ignores user acquisition & conversion. Freemium often means you "take a paycut" on option A in order to make even more money on option B.

> Comparing 12.5% revenues to a pay cut is not misleading at all.

It's misleading in the sense that a 12.5% pay cut is not something I can recover in other ways.

Meanwhile, there is a possibility that this 12.5% cut on individual account might be recovered automatically in other ways if it results in increased enterprise account.

My personal private repos have always been on BitBucket. Not because I like BitBucket more, but simply because of the additional cost.

Having unlimited private account on GitHub will encourage me to move my stuff there, and ultimately reduces the friction and increases the chance that at some time in the future I might be a paying enterprise customers.

If we’re going to use analogies like that, then this is like my benefactor uncle coming in with massive sums of money, telling me to stop working at the McDonald’s and pay someone to take my shifts, and focus on studying.

Yeah, I take that deal every day of the week.

There's maximum probably only like hundreds of thousands of people who would pay $96/yr for git repos for their personal projects. Millions a month but their other revenue streams are more expensive and still growing and can keep growing so it was heading towards a rounding error % even though it was 12% a couple years ago.

I imagine they are making predictions of enterprise business influx through new users who will take their personal private repos and possibly business (back) to github.

It's probably shrinking dramatically on a percentage basis -- Microsoft's massive enterprise presence drive tens or hundreds of thousands of accounts.

My business may be able to spend 12% of revenue on advertisement.

>Your assertion can only make sense if you also believe that a 12% paycut is irrelevant to your income.

Beyond a certain hand-to-mouth income it is (e.g. 150K vs 132K).

And many people chose pay cuts over personally overworking or stretching their business too thin.

The overly trivial point being behind the parent argument is that 100% of your revenue if better than less than 100%. Which is not always the case in business.

Businesses and freelancer drop gigs all the time to concentrate on where their main income is coming from. If you have a difficult customer or few demanding customers that generate just 20% of your income, it's often wise to drop them to focus on your easier 80%.

Your assertion can only make sense if the same economic factors and reasoning that govern a private individual's pay also govern a large multinational organization's revenue.

> Your assertion can only make sense if you also believe that a 12% paycut is irrelevant to your income.

You only see the short term.

I was on Github during my university, all my projects were there, my team mate were on there too, it was pretty great. When my student account expired, I decided to move everything to Gitlab. I didn't want to move but I won't pay 7$ a month to host my old school projects and some personal stuff. Now that I did move everything, I actually enjoy Gitlab quite a bit.

What do you think I'm going to use if I start a commercial project? In the past I would have chosen Github, but now that I'm comfortable with Gitlab and see all its benefits, it's probably the one I would choose.

Cutting that 12% may lose 12% in the short term, but in a few years, it's going to gain you much more than 12%.

I totally agree with you. And to add to your point, the impact of having a dominance in the market of "personal project repos" extends far beyond individual developers paying money on their personal accounts and/or becoming paying customers when they start their own companies.

If you think about mid-sized software companies (~10-100 devs), if 80% of your developers have a strong preference for Github over Gitlab or Bitbucket, then that company is likely going to stick with Github or decide to move to Github eventually. I think this is where Github captures much more than that 12% revenue that they are giving up right now.

Also, if the VP of Eng. or whoever is making the decision thinks of Github as the default git SaaS because it's what they have used since college, there's a good chance they will just go with Github without even evaluating alternatives.

Yep, there's a reason companies like Autodesk and Microsoft et al give students thousands of dollars of software for free while they're students.

Glad you appreciate the GitLab experience! I agree with you that loving the product is what it is about. I am happy everyone gets free repos across the board now (whether at GitLab, GitHub, or BitBucket) since repos are basically commoditized at this point. In my opinion (shaped by our CEO Sid's perspective), all roads lead to compute.

I can say the same, except in my case it was moving from Github to Bitbucket. If Github, did this a while back I would have been locked in for life (by choice)

+1 on that

I moved all my personal repos to Bitbucket a while ago, as it was free, and have referred them several paying customers from my clients.

Sure, but that 12% isn't irrelevant, it's an investment. They're investing 12% of lost revenue in future growth.

Now, if personal accounts were far less (e.g. 1% of revenue), I could see the argument that they're irrelevant.

Maybe I'm just complaining about vocabulary, and perhaps your clarification was what the OP intended.

That was in 2016 it is likely to be a lot lower now, as enterprise grew.

It's a tradeoff. I took a 11% paycut to become a 100% remote worker. Completely worth it for me. Microsoft likewise likely believes the loss of 12% of GitHub's revenue is worth it in exchange for advancing toward their bigger goals.

You're missing why Microsoft purchased github in the first place. It wasn't for the personal account revenue - which isn't even a rounding error on their earnings report even if it is 10x what it was when revenue was leaked in 2016 (at which point it was $17.5 million/year).

If that WERE the reason, they'd be charging for visual studio code, not giving it away.

GitHub's competitor is GitLab. The main argument in favor of GitLab is private repo with self hosted GitLab. That's a weak spot of GitHub that they addressed. Self hosting requires a server and competence. With this move they also undercut a revenue source of GitLab. Strategically, it's a brilliant move. From the user point of view it's also an excellent news.

In what world is 12.5% of revenue "irrelevant"? The 37.5% business revenue is just 3 times as much (so almost as irrelevant), and Enterprise is only 4 times is much (again, almost as irrelevant).

Unless .. all github revenue is irrelevant for Microsoft - which might be true, for all I know.

I bet it's much less than 12.5% now (since 2016). Also, individual user revenue is revenue that public investors value less than revenue from businesses, because there is perceived to be a much bigger overall market (and therefore more future potential growth) in selling to businesses rather than individuals. And growth is what matters for the stock price.

Source: I am the CEO of Sourcegraph and consider similar things.

What I've always found funny about "B2B is more valuable than B2C" is how all economic value eventually comes from B2C eventually, and all the meta B2B tools in between don't really make value independently on their own.

Yes, but all consumers eventually get their money from companies. So....

But in reality, the source of your customers' income has little to do with value. B2B is favored because the margins are generally higher.

>In what world is 12.5% of revenue "irrelevant"?

Meh, just fire 12.5% of the staff. /s

That might indeed be true. GitHub's value to Microsoft ain't the revenue; it's the users, the mindshare of whom Microsoft hopes to capture by luring them toward the Microsoft ecosystem (Windows, Office, Azure, SQL Server, etc.).

As a particular former Microsoft CEO put it: "[…] DEVELOPERS DEVELOPERS DEVELOPERS DEVELOPERS […]". That strategy/desire hasn't exactly stopped, though it's certainly transformed quite a bit over the last couple decades (and Microsoft is finally actually getting somewhat effective there).

12.5% is still a significant percent of revenue.

Significant... but the main threat to the other 87.5% is that developers try out Gitlab or other alternatives for their personal projects. These developers find that Gitlab is pretty good, and don't push their companies to pay for Github. Whereas if every developer starts out using Github exclusively for personal projects, it becomes the default choice when they start work.

With this change, Github foregoes a little revenue in exchange for securing the vast majority of their revenue.

Exactly this. Just this weekend, I setup my first CodeCommit (AWS) repo because I needed a private git repo for a new project and didn't want to pay github for it.

It turned out to be damned easy (and very cheap) and that's a serious threat to github's future growth because once people start to default elsewhere for one type of their work, it tends to drag along other types of their work.

That was always the case, though. It's just that Github's network effects were always powerful enough to counteract this effect.

But, for one very obvious reason, people have grown more and more interested in finding alternatives to Github, and they're forced to compete.

Unless it's shrinking as devs realize other platforms offer a free version of the same service. Especially if those devs decide that they like the new platform better, and convince their company to use it.

It is for a lot of companies, but not at the scale of Microsoft. Keep in mind companies like Google/MS/Apple/etc kill products that have 10s of millions of revenue because while it's a ton of money, it's a drop in the bucket compared to their other sources of revenue. 12.5% of their 300M revenue is a tiny amount for MS to absorb to solidify themselves against their competitors as the main open source hub.

Gitlab and Bitbucket both heavily positioned themselves against Github with their unlimited private repos feature. With this competing with Github is going to become much harder. I think it's a smart move for MS.

Reply on news.ycombinator.com