What Does Cloud-First, Mobile-First Mean for Microsoft?

I’ve been meaning to write about this for some time now but with all the tech news lately I’ve been focused on other things. In Satya Nadella’s first email to employees on the day he took over as CEO, he said, “Our job is to ensure that Microsoft thrives in a mobile and cloud-first world.” That mobile-first, cloud-first refrain has been repeated several times since, and it was even used as the title of the press conference in May where Microsoft unveiled Office for iPad, among other things. So — what does it mean? What does it mean for Microsoft?

I think first of all there’s been some confusion about how both mobile and cloud can be “first” – Nadella even alluded to this in the aforementioned press conference to laughs from the audience:

Now, we describe this, and I talked about this in my first mail to all Microsoft employees as a mobile-first, cloud-first world. And like any place that’s got predominantly computer scientists, I get back this mail saying, hey, look, how can two things be first? I mean, do you have a problem with ordinal numbers or something?

The answer, of course, is mobile and cloud sit in different spaces in the overall scheme of what Microsoft is trying to do, as shown in the graphic below:

“Mobile” refers to endpoints. Your definition of mobile might be narrow, referring only to what we used to call mobile phones; it could be slightly broader, referring to all those devices we primarily hold in our hands rather than setting them on a surface, such as tablets; or it could mean all those devices which are in a fundamental sense portable, including laptops.

Regardless of which definition you use, mobile is about the endpoints and devices on which we consume (and create) content, engage in communications, and generally get things done. Cloud, on the other hand, is, in a loose sense, a location and a philosophy: an architectural approach to where data and intelligence should live as part of an overall product or service. Whether or not this data or intelligence sits in a distributed cloud environment is completely separate from the endpoint it happens to be consumed on. To be cloud-first is to design new products and services with cloud back-ends in mind and to be mobile-first is to design them with mobile front-ends in mind. Those two complement each other well, and there’s no conflict here.

The other major difference between the cloud and mobile aspects of Microsoft’s strategy is decisions about these things are made by different people: individuals primarily choose their endpoints and care deeply about which they use. But organizations primarily choose the location of the back-end data and infrastructure, and if products and services are designed and architected well, end-users should be blissfully unaware of that architecture. That applies both to enterprises deploying services used by their employees and to developers creating services for consumers. Think of Gmail, arguably one of the biggest mass-market cloud services out there: do any significant number of its users think of it as a cloud service, or wonder where their email is stored? Of course not. They do care about whether it’s available and works well on their devices, and that may even be a factor in which devices they choose to use, but the fact it’s a cloud service is neither here nor there. It is arguable, of course, that mobile endpoints, with their limited storage and computing capacity and their tendency to move around a lot, are a perfect match for cloud back ends, which tend to be distributed and can readily serve up relevant data and functionality to mobile devices on a reliable and ubiquitous basis. But again, that’s not something end users should have to worry about.

Being cloud-first means designing new products and services to be delivered from a cloud infrastructure rather than sitting on a server in a company office or on a computer in a home. That has significant implications for how services are designed and delivered, as any enterprise looking to migrate legacy systems from on-premise to cloud environments has discovered. Systems designed for on-premise models don’t take into account the latency introduced when servers and endpoints are separated by hundreds or even thousands of miles, and tend to break down pretty quickly when they’re moved to cloud environments. But systems designed with cloud environments in mind can easily be deployed in an on-premise fashion, so this is a sensible move on Microsoft’s part.

Microsoft is also a primary provider of cloud services, both for consumers and for enterprises, and that part of its business is expanding rapidly. There’s been a lot of debate about “who has the biggest cloud” this past week, but the answer is somewhat irrelevant: Google, Microsoft and Amazon are three of the largest providers of a certain class of enterprise cloud services, and that’s all that matters. The combination of consumer-facing cloud services such as Outlook.com, large-scale enterprise cloud infrastructure services such as Azure, and a shift to new, cloud-based business models such as Office 365 for consumers and business end users is a major strength for Microsoft as cloud models become more important in the overall architecture of modern technology. Microsoft is well placed to succeed in all these categories and controls much of the underlying infrastructure needed to deliver it.

In mobile on the other hand, even if we take the broadest possible definition which includes laptops, Microsoft has a tiny share of the devices and a minority share of the platforms which run on these endpoints. Even though Nadella apparently has plans to grow beyond what Microsoft has described as a 14% share of computing devices, the reality is Microsoft’s products and services in a mobile environment will primarily be running on others’ devices and platforms. That, in turn, means Microsoft’s mobile-first focus will be implemented primarily in the user interfaces for individual products and services. But this is where your definition of mobile becomes really important: for certain of Microsoft’s products, starting with the smartphone makes perfect sense: Skype, for instance. But for others – Office being an obvious example – focusing on a device where most of the use cases will likely never be employed doesn’t make sense. Tablets and PCs should be the focus for Office, with a subset of the features available there available on smartphones.

The bigger challenge, though, is Microsoft will increasingly be competing against the owners of the two major platforms even as they seek to broaden their offerings onto these platforms. Apple and Google both have their own offerings which compete with Microsoft’s Office, Skype, OneDrive and Outlook.com products. Their competing products are in most cases free to the user, bundled into a broader suite of online services or into a hardware purchase. How will Microsoft compete against these two companies on their own platforms in a way that both adds enough value to justify charging money and overcomes the disadvantages of being a second class citizen? Other commentators are saying Microsoft should abandon both its devices and platforms businesses and focus on cross-platform services, but the fact Microsoft’s two major competitors own those platforms should be cause for reflection.

In summary, there’s no conflict, as such, between being cloud-first and mobile-first: one refers to architectural decisions and the other to endpoints and user interfaces, and the two are complementary. Organizations care deeply about the former, while end users care deeply about the latter. The bigger challenge for Microsoft is, although it is increasingly strong on the cloud side, it remains weak on the mobile side, and will be trying to compete on platforms owned by its two major competitors. I continue to be skeptical it can succeed in making money from its current set of products and services as long as they’re primarily running on Apple and Google’s platforms.