Is Microsoft’s Windows cloud strategy a case of Azure-based frog boiling?

Microsoft has said that they’re a “mobile-first, cloud-first” company these days. Windows 10 represents the first wave of the new world, and already there are signs that a cloud strategy lies firmly underneath some of the more questionable aspects of the operating system.

There are approximately five billion desktops in the world, and getting even a small percentage of these into a cloud-based system would represent an astonishing amount of monthly revenue. But in order to make hosting, and managing, this kind of volume of systems even remotely acceptable, you need to reduce the complexity as much as possible.

Microsoft consultants, when they’re not repeatedly trying to re-engineer solutions in response to simple questions, tend to try and switch people to more simplistic models of delivery. One device per user. Local profiles. Small amounts of roaming settings provided by simple XML files. Device-based customizations. Given that we used to be able to manage all of these things on a very specific basis, Microsoft’s new stance is somewhat surprising, but, in my opinion, is possibly indicative of a wider strategic play.

Take, for instance, settings like internet home pages and file type associations. Microsoft has changed the policies for these settings to specifically device-based, and more are steadily following the same curve. This means if you have Remote Desktop Session Host servers or shared desktop devices, you can’t manage the provision of these settings to your environment on a user-specific basis. Even using user environment management tools is made difficult by a plethora of device- and user-specific hashes within some of the values used to provide these settings. Microsoft has pointed rather shakily towards a security standpoint for these adjustments, but I can’t help but feel there’s more to it, and that pushing these kinds of simplifications makes it easier for them to achieve a wider goal.

Roaming of user settings, and multi-user systems with lots of different customizations, bring both management and administrative complexity. By changing policy objects that deal with these particular settings, Microsoft is – in the short term – generating an amount of annoyance amongst support teams, but what do they potentially gain from it? What can they possibly achieve?

By handing back a degree of control to the user (as in this example, for FTAs and home pages), Microsoft is reducing the administrative complexity of traditionally quite convoluted systems. But what they also do is make these systems, once they are bedded in, much more desirable targets for migration to their cloud strategy Azure-hosted model. Currently, client-side instances of operating systems are not permitted in Azure, but it only seems a matter of time before they become offered – and when this happens, Microsoft, who will be responsible for them, will need to make sure they can be managed in the most cost-effective (read: cheap!) way possible.

Cloud strategy?  Or ‘boiling a frog’.

It’s rather like boiling a frog – I’m sure you’ve all heard the analogy. Apparently, if you pop a frog into a pan of boiling water it will leap straight out, but pop it in cool water and heat it very gently, and eventually, it will boil to death. Microsoft are encountering initial push-back over the change to the management model, but currently, administrators are finding ways around this, either by scripting or using third-party software. But the more they become used to it – as the water steadily heats up – then eventually they become ripe targets for the carrot of Azure migration, and the potential cost savings that Microsoft sales people can tempt them with. And from then, of course, with Windows 10’s fast update cadence, they can let you pay for the management, serve you the ads, and ramp up the data mining.

Think it’s unlikely? Look how much Office 365 is taking off as another cloud strategy. Not the “full web” version, but the version that still uses the Office client – “Exchange Online”, if you prefer. Is this the future model, or is it just that everyone hates managing Exchange? Once the issues around Cached Exchange Mode are sorted out – and there are features in the pipeline that can address this – then Exchange Online will become possibly the de facto model for email provision. And if Exchange can be managed in this way, you can bet your life Microsoft are going to try and bring more lines under the same Azure-hosted platform.

Of course, there are lots of bumps in this road. Industry vertical, compliance, security, data sovereignty, general paranoia – there are a lot of factors to take into account. But, like the poor frog, once the water starts heating up, there often isn’t a way to turn it down – unless you want to make a big jump away from Microsoft’s ecosystem.