If I have a major takeaway from the Amazon Web Services re:Invent conference this week (and Google Cloud’s parallel news stream), it’s that we’re witnessing the moment when cloud computing evolves into mainstream, table-stakes enterprise IT. You can see this not only in the sheer size and attendance of the the AWS event, or the Fortune rankings of the companies highlighted as customers, but also in the increasing homogeneity of the AWS, Microsoft Azure and Google Cloud platforms. As competition among those three pick up, there’s a pressure to fill perceived gaps that might make it more difficult to figure out who’s who.
When Google entered the cloud market in earnest, the conventional critique was that it has cutting-edge technology but no experience working with large enterprises. Now, with Bryant and former VMware CEO Diane Greene at the helm, you can’t really make that claim – at least at the leadership level. Google still has some very smart technologists and boundary-pushing technologies, but the company is pretty clearly trying to send a message that it must be taken seriously as an enterprise IT provider on par with AWS and even Microsoft.
If anything, Microsoft is probably the party responsible for pushing its peers in this direction as Azure grew into a viable option. AWS had to – and certainly did – react to this new threat to its market dominance, and Google had no real choice but to do the same if wanted to be taken seriously.
On the flip side, however, it was easy to see Google’s influence in the
spate of announcements made by AWS this week. The Google-created Kubernetes technology is obviously responsible for pushing AWS to create its own Kubernetes service for container orchestration, and almost certainly inspired the “serverless” Fargate orchestration service – if only as a way to highlight how developer life could look beyond the complexity of Kubernetes. Microsoft is
also a major player in the Kubernetes world.
And artificial intelligence. All the AI services AWS announced this week, from SageMaker (essentially a platform for building, training and optimizing AI models) to the various voice/text/video recognition services, owe their existence to Google. Not just because Google essentially kicked off the current deep-learning-based AI obsession back in 2012-2014, but also because Google Cloud offers its own versions of these things. So does Microsoft.
All of this competitive pressure is great for customers and probably will be great for cloud providers in the short run, but in the long run it seems like it will only
really be great for AWS. Not only does homogeneity favor the incumbent, but AWS is also doing a really good job pushing the issue and distinguishing itself around serverless computing. That started with Lambda, but now has
advanced to its Aurora database service and (I believe) some other places, well.
Competing on the merits of your respective products is a noble strategy, but it’s not always going to be enough. You can’t blanket the world in your cloud platform the same way you can blanket the world in Android phones to put the pressure on Apple. We’ll have to see how Google and Microsoft ultimately distinguish themselves from each other and from AWS in a meaningful and lasting way, while still making sure they’ve got all the must-have boxes checked.