Why does Microsoft care about GitHub to the tune of $7.5bn? I suspect there are two primary reasons:
- Adding to the Microsoft Graph, their ever-growing repository of interconnected insights into identity, organizational membership and activity. Back in the day, Active Directory was the asset that made Windows indispensable in businesses. Graph extends the stickiness of Active Directory to the cloud, which is why it’s such a precious asset to Microsoft. Microsoft would love to be able to link developers with organizational identity, skills profiles, open projects, training needs, APIs and so on; and Office 365, LinkedIn and GitHub together provide a treasure trove of metadata to mine and monetize.
- Secondly, as ever it’s all about the cloud. If GitHub source is hosted on Azure with built-in CI/CD, and a GitHub account with attached payment instruments is one click from an Azure deployment, then you can see plenty of upside in driving Azure adoption among the single audience that can make or break the cloud.
It’s easy to see how Microsoft can gain from a transaction like this, even at an inflated premium — and just as importantly, how threatening to Microsoft it would be to see a competitor acquire GitHub. I’m sure that Microsoft already had gamed this out; the moment it was clear that GitHub was open to a sale it seems that Microsoft jumped before they could be locked out.
There haven’t been many successful acquisitions of this scale at Microsoft. Nokia ($7.2bn) was perhaps the most prominent example of squandering shareholder value, but there have been others. Skype ($8.5bn) isn’t yet described as a failure, but it seems evident that Microsoft have thus far failed to unlock the value that they saw in the deal, particularly in the consumer space. Yammer ($1.2bn) was essentially a defensive purchase that protected Microsoft’s enterprise collaboration interests, but it’s again not clear that it was worth the disruption or the confusing mess of competing products that Microsoft has to offer in this space as a result. Their largest single acquisition, LinkedIn, seems more successful, although it will likely take many years yet to be certain that they have added more than the $26bn paid to their asset value.
Perhaps the one that is most instructive is Microsoft’s disastrous acquisition of aQuantive, which led to a $6.2bn writedown in 2012. In a Geekwire article, aQuantive execs call out a few reasons for failure that are useful learnings for the GitHub acquisition team:
An inability to let aQuantive operate as a fast-moving standalone unit. A long, drawn-out brain drain of aQuantive’s top minds. Too many disparate aQuantive businesses to integrate. But most of all, the former aQuantive staffers say it was simply a culture clash.
The first of these challenges seems particularly pertinent. There’s an implicit tension in an acquisition like this. On the one hand, Microsoft will I’m sure tout that it’s “business as usual” and that GitHub will continue to operate much as it has done. But that doesn’t justify $7.5bn; the value only comes when you take advantage of the network effects of combining GitHub and other Microsoft properties, which is inherently somewhat customer-hostile (at least to some percentage of the GitHub user base).
And what about the ailing Visual Studio Team Services (VSTS)? Will Microsoft attempt to merge these two products, with all the disruption that causes to customers, employees and strategy? I imagine a long drawn-out death for that product, as they attempt to knit various parts of GitHub and VSTS together. What organization would now commence a significant VSTS deployment, knowing that Microsoft had just bought GitHub?
What matters next is how Microsoft executes and how the community will receive this news. Will the initial rumblings from GitHub customers moving to GitLab turn into an avalanche? Can Microsoft persuade GitHub customers that they have their best interests at heart? Will the open source community feel comfortable to host their public source code with Microsoft? Will the gradual convergence of open source onto GitHub come to an end, with competitors like GitLab and Bitbucket benefiting greatly? These are not questions that can yet be answered by Monday morning quarterbacks.