Venture capitalist Brad Feld recently wrote an interesting post predicting the end of Amazon’s dominance of the cloud computing market, and concluded, “it’s suddenly a good time to be Microsoft or Google in the cloud computing wars.”
I’d go one step farther. Using Feld’s arguments, I’d say that Microsoft is in the driver’s seat.
First, the price war. Microsoft and Google are on approximately equal ground when it comes to cutting prices — both have highly profitable core businesses that they can use to subsidize a price war in cloud infrastructure, even to the point of sustaining losses for a while to gain market share. Amazon does not.
Second, the quality argument. Like Feld, we’ve also pointed out that there are niche cloud providers that do a better job than the big guys at providing infrastructure-as-a-service for specific verticals, but when you move all the way up the stack to full software-as-a-service applications, Microsoft has an edge among the big three with Office 365. Google has been making inroads into smaller businesses with Google Apps for almost a decade now, Microsoft remains the standard in the biggest and most profitable business customers — as this recent investigation from Dan Frommer at Quartz showed, only one company in the Fortune 50 uses Google Apps. (That company happens to be Google itself.)
The third argument, support, is mostly a wash. While Amazon’s support may be terrible (I have no evidence of this, but I’m taking Feld’s word for it), Microsoft and Google and their respective ecosystem partners do a decent job of supporting customers on their stacks. This hasn’t always been the case — Google used to treat support as an expensive afterthought — but in the case of Google Apps, at least, the company and its partners have stepped up significantly.
But then comes the fourth argument. Feld points out that once companies get to $200,000 per month of cloud-infrastructure spend, it’s actually significantly cheaper to build their own data centers.
Microsoft is the only one of the big three players with an on-premise offering — Windows Server and the rest of the Microsoft infrastructure family. Maybe the exact break-even point will change as the cloud price wars continue, but Microsoft has the most pieces customers would need to move from all-cloud to a hybrid or on-premise solution. Or, for that matter, for existing on-premise customers to begin experimenting moving some workloads to the cloud.
There’s one more point favoring Microsoft. Google’s core business is selling online advertising. That business makes up about 90% of Google’s revenue, and it has enviably high operating margins — around 30%, based on Google’s 2011 financial report. (I picked 2011 because that was before Google bought Motorola Mobility, which changed the margin structure.)
It’s unclear how the Google Cloud Engine helps that business. Are customers using Google’s cloud somehow more likely to advertise with Google? I don’t see it. Are Google advertising customers demanding to run other workloads on Google technology? I don’t see it.