A few days ago, Microsoft published The Economics of the Cloud, a whitepaper that has so far not gotten nearly as much attention or consideration as it deserves.  Perhaps this indifference is due to a collective freshman flashback on the dreaded "Econ 101" or, to skepticism about Microsoft's importance in the new world of cloud computing.  Either way, it is unfortunate because the paper presents some startling new data about the cloud, and, not entirely intentionally, reveals the company's cloud strategy at a level of nuance that we have not seen before.

The paper is by Rolf Harms and Michael Yamartino, a director and manager, respectively, in Microsoft's Corporate Strategy Group.  The paper is pointed at "IT leaders", a phrase used 16 times in 22 pages, and has the ostensible goals of sharing the cloud wisdom Microsoft has gained from doing Azure, Bing, Windows Live, and Office 365, and of sharing data gathered and conclusions drawn by Microsoft about the future of cloud computing from modeling done in its Strategy Department.

The numerical data in the report and what the researchers make of it are quite interesting in their own right, but, when calibrated, Da Vinci Code style with the company's history and recent activities, they may reveal a bit more than the authors intended.  In any case, neither Ray Ozzie's dreamy "Dawn of a New Day" farewell memo nor Steve Ballmer's buzzy "All In" UW speech and internal memo helped us see this coming.

I admit it, until this whitepaper, I was one of those who took the "All In" stuff for opportunistic hyperbole and simply did not believe Ballmer when he said in his memo, "We need to be (and are) willing to change our business models to take advantage of the cloud."  Yeah, given what the cloud will do to the licensed software business, he should be saying that.  But, a company that big and successful changing its business model seemed impossible.  In short, I thought the cloud would turn Microsoft into the world's biggest dairy farm.  Now, I am not so sure.

Biggest is Bestest

The intended takeaways from the paper are summarized like this:

"Private clouds address many of the concerns IT leaders have about cloud computing, and so they may be perfectly suited for certain situations.   But because of their limited ability to take advantage of demand-side economies of scale and multi-tenancy, we believe that private clouds may one day carry a cost that is as much as 10x the cost of public clouds."

"Based on our analysis, we see a long-term shift to cloud driven by three important economies of scale: (1) larger datacenters can deploy computational resources at significantly lower cost than smaller ones; (2) demand pooling improves the utilization of these resources, especially in public clouds; and (3) multi-tenancy lowers application maintenance labor costs for large public clouds. Finally, the cloud offers unparalleled levels of elasticity and agility that will enable exciting new solutions and applications."

They are saying that the private cloud will be a niche business for them and a costly specialty for customers - the future is all about big public clouds, due to their dramatic economies of scale gained through lower infrastructure costs, higher utilization, and multi-tenancy cost amortization.

Microsoft must be changing to a new business model, because those things are all bad for their old one.  Today they make most of their money from dedicated servers, desktop software, and single-user and single-tenant applications.

Supply, Demand, and Multi-Tenancy

The supply-side economies of scale gained through big public clouds highlighted in the paper are:

• Lower power costs through strategic power grid location and bulk purchasing
• Lower labor costs from fewer employees managing more servers and apps.
• Higher security and reliability due to provider expertise and infrastructure quality
• Higher buying power from high volumes of a few standardized configurations

The demand-side economies of scale are gained through optimizing infrastructure utilization in these five areas:

• Randomness of end-user access
• Time of day patterns for applications
• Industry-specific variability
• Multi-resource variability
• Uncertain growth patterns
• About these factors, the paper says,

"A key economic advantage of the cloud is its ability to address variability in resource utilization brought on by these factors. By pooling resources, variability is diversified away, evening out utilization patterns. The larger the pool of resources, the smoother the aggregate demand profile, the higher the overall utilization rate, and the cheaper and more efficiently the IT organization can meet its end-user demands."

In other words, the bigger the cloud, the more the users, and the more diverse the applications, the greater the economies of scale on the demand (customer/user) side will be.

Finally, the report breaks out the multi-tenancy economies of scale like this:

• Fixed application labor amortized over a large number of customers.
• Fixed component of server utilization amortized over large number of customers.

The whitepaper elaborates on these factors in great detail to make a compelling case for big clouds and then goes on at length to impugn the private cloud and provide reassurance that the common IT concerns of security and compliance about the public cloud would soon be non-issues.

The report's big finish is an analysis of two points of probable market and the conclusions that should be drawn by "IT leaders."

We can do this the easy way or the hard way.

The first behavioral point is that "Decentralized IT (also known as ‘rogue IT') will continue to lead the charge."  In support of this as follows:

"Many prior technology transitions were led not by CIOs but by departments, business decision makers, developers, and end users - often in spite of the objections of CIOs. For example, both PCs and servers were initially adopted by end users and departments before they were officially embraced by corporate IT policies. [...]

"We‘re seeing a similar pattern in the cloud: developers and departments have started using cloud services, often without the knowledge of the IT group (hence the name -rogue clouds‖). Many business users will not wait for their IT group to provide them with a private cloud; for these users, productivity and convenience often trump policy. [...]

"CIOs should acknowledge that these behaviors are commonplace early in a disruption and either rapidly develop and implement a private cloud with the same capabilities or adopt policies which incorporate some of this behavior, where appropriate, in IT standards."

In other words, Hey, CIO, get with the program or we will help the business users go rogue on you.

Go public.  You'll feel better.

The second behavioral point is that "Perceptions are rapidly changing," supported by the observation continued SaaS successes are building trust in the cloud that will only increase, and closed out with this carrot and stick:

"In summary, while there are real hurdles to cloud adoption today, these will likely diminish over time. While new, unforeseen hurdles to public cloud adoption may appear, the public cloud economic advantage will grow stronger with time as cloud providers unlock the benefits of economics we discussed [earlier.]  While the desire for a private cloud is mostly driven by security and compliance concerns around existing workloads, the cost effectiveness and agility of the public cloud will enable new workloads."

In the final section, the paper, this last point is reinforced in this way:

"For businesses of all sizes, the cloud represents tremendous opportunity. It represents an opportunity to break out of the longstanding tradition of IT professionals spending 80 percent of their time and budget -keeping the lights on, with few resources left to focus on innovation. Cloud services will enable IT groups to focus more on innovation while leaving non-differentiating activities to reliable and cost-effective providers. Cloud services will enable IT leaders to offer new solutions that were previously seen as either cost prohibitive or too difficult to implement. This is especially true of cloud platforms (Platform as a Service), which significantly reduce the time and complexity of building new apps that take advantage of all the benefits of the cloud."

In other words, Why don't you IT people leave all that infrastructure management to someone else and do something new and innovative?

What are they really saying?

They seem to be trying to talk IT out of implementing private clouds, or any other kind of internal IT infrastructures, so that means that Microsoft is moving away from selling stuff directly to IT.  But, if the public cloud is the future, does that mean that their new customers will be cloud service providers, like telcos or Rackspace, or Salesforce.com?  Maybe not.

It appears to me that the end-game for Microsoft is to become the biggest public cloud in the world, comprising AZURE PaaS, plus Office, Bing, and Windows Live SaaS, and partner applications re-engineered as multi-tenant services.  They put it like this:

"We have over 600,000 partners in more than 200 countries servicing millions of businesses. We are already collaborating with thousands of our partners on the cloud transition. Together we are building the most secure, reliable, scalable, available, cloud in the world. [...]

"Microsoft and our partners helped bring PCs to over one billion homes and desktops. Millions of developers and businesses make their living on PCs and we are fortunate to play a role in that. [...]

"Now, we have a vision of bringing the power of cloud computing to every home, every office, and every mobile device. The powerful economics of cloud drive all of us towards this vision. Join Microsoft and our partners on the journey to bring this vision to life."

So, their customers remain the same -SME and enterprise IT.  Their functional products remain the same - applications and enablers for custom and partner applications.  What changes is the way that functionality is delivered to those customers, with a large portion of private IT and its supporting vendors evaporating and transferring value and power to Microsoft and their really big honkin' cloud.

In the whole carefully written whitepaper, there was not a single mention of other cloud service providers and how Microsoft will partner with them, because they may not.  It is conceivable that Microsoft would license Azure to, say, hosting services currently based on Windows, but I think it is unlikely.  There is no reason to think that Microsoft doesn't believe that the economics are telling them that they can dominate the new world as much or more than they did the previous one.

• Will Office 365 Redefine Enterprise Applications?