Ben Thompson is generally spot-on in his analysis of industry goings-on. But he missed a lot in The Cost of Developers this week.
Here’s what he got right about this acquisition:
- Developers can be quite expensive (though, $7.5B (in equity) is only ~$265 per user (which is pretty cheap))
- Microsoft is betting that a future of open-source, cloud-based applications that exist independent of platforms will be a large-and-increasing share of the future
- That there is room in that future for a company to win by offering a superior user experience for developers directly, not simply exerting leverage on them
- Microsoft is the best possible acquirer for GitHub
- GitHub, having raised $350 million in venture capital, was not going to make it as an independent entity
- Purely enterprise-focused companies like IBM or Oracle would be tempted to wring every possible bit of profit out of the company
- What Microsoft wants is much fuzzier: it wants to be The Developer’s Friend™
- [Microsoft] will be ever more invested in a world with no gatekeepers, where developer tools and clouds win by being better on the merits, not by being able to leverage users
And here’s what he missed and/or got wrong:
- [Microsoft] is in second place in the cloud. Moreover, that second place is largely predicated on shepherding existing corporate customers to cloud computing; it is not clear why any new company – or developer – would choose Microsoft
- It is very hard to imagine GitHub ever generating the sort of revenue that justifies this purchase price
Some of the below I commented on Google+ yesterday. The rest is in response to more idiocy & paranoia I’ve seen on some technical community mailing lists (bet you didn’t know those still existed) in the last 24 hours, or in response to specific items in Ben’s essay that are shortsighted, misguided, or incredibly wrong.
- If you cannot see why new users, developers, and companies would go to Microsoft Azure offerings, you don’t understand what they’re doing
- AWS is huge – but Azure and Google Cloud Platform (GCP) have huge technical (and economic) advantages
- Amazon likes to throw new cloud features at the wall like spaghetti to see what sticks; Google and Microsoft have clearly thought through this whole cloud business, and make incredibly solid business & technical sense to use over AWS in most use cases (the only [occasional] real exception being “but we already use AWS”). Have you not seen the Azure IoT offerings?
- GitHub has not yet been profitable, and would probably have IPO’d (poorly) in the next year to keep from running out of cash
- Arguably, GitHub would never become profitable on their own
- Microsoft has a long history of contributing to OSS projects (most-to-all of which are on GitHub)
- If they were going to acquire anyone in this space, GitHub is the only one that makes any sense
- (This was tangentially-mentioned in Ben’s essay by linking to his analysis of the Microsoft-LinkedIn acquisition in 2016.) Alongside the LinkedIn acquisition a couple years back (which has an obvious play for an eventual IDaaS (fully-and-forever integration with Office365 regardless of where you work, everything follows automagically)), offering better integrations with their existing tools (Visual Studio already had git integrations – they should only get better with this acquisition) is a Good Thing™ for devs and end users alike (because making those excellent developer tools even better means they’ll be better whether they’re using GitHub, Bitbucket, GitLab, etc)
- The more-or-less instantaneous expansion of offered items in the Windows Store (some kind of cloud-based/distributed build-on-demand for software when you want it (and which fork you want)) to “everything” on GitHub is a brilliant possibility
- In light of Apple’s announcement yesterday about enabling iOS apps to come to macOS over the next releases of iOS and macOS, this should have been at the forefront of most people’s thought processes (after the keynote was done, of course)
- Through this acquisition, it’s [probably] likely more developers will use Microsoft APIs (.NET, etc) in their projects
- Echoing Ballmer’s chant, “Developers! Developers! Developers!”, while Microsoft doesn’t really care about Windows anymore (just look at the recent reorg), it is still THE most widespread end-user platform in the world – and bringing millions more developers “into the fold” is genius
- Even if some small percentage will opt to go elsewhere, most won’t change because, well, change is hard
- All the developers Microsoft had that weren’t yet using GitHub will have a huge reason to start
- Microsoft has typically been a buy-don’t-build shop (there are exceptions, but look at the original DOS, PowerPoint, SQL Server, Skype, their failed attempt at Yahoo!, etc): they could have spent 5-10x as much building something “as good as” GitHub, or they could buy it; they opted for the “buy” (via equity, note, and not cash (smart from several business viewpoints (not least of which is the “enforced” interest the GitHub subsidiary (with its new CEO, etc) will have in continuing to ensure it is The place for developers to put their projects (after all, if that drops considerably, the equity aspect GitHub got in the deal is going to drop))))