An interesting debate took place at last week’s Cloud2020 gathering regarding the viability of futures markets for cloud computing capacity. I’m firmly at the skeptical end of the spectrum as the title of this post will attest. However, I had not given enough thought as to the reasons for my skepticism. Having reflected on it a little I’m more convinced than ever that any attempt to create either primary or secondary markets in cloud commodities is doomed. In short the lack of friction and a lack of volatility in matching cloud computing supply and demand means there is likely no window for 3rd party market makers to insert themselves into this value network.

I am neither and economist nor do I have any experience in the financial markets. At best I’m an informed observer. My opinions are therefore not based on any particular expertise in this domain. That withstanding my views are based on my observation that the basic nature of cloud computing production, delivery and consumption is unlikely create the opportunity for the emergence of primary and derivative markets.

Primary Markets

A ‘Market’ in its most simple form is a mechanism for matching buyers and sellers of goods, services or information. The market acts as a means of matching supply and demand and therefore arriving at the optimal price at which goods will be exchanged. The more buyers and sellers there are active in the market — allowing for perfect sharing of information across the market — then the closer the market will get to providing the optimal exchange of value between buyers and sellers.

More formally; ‘Markets’ create local minima in complex trading networks that minimize the amount of energy that needs to be expended to trade by maximizing the amount of information available to market participants. This is the reason we have fish markets.

Consumers of fish do not have direct access to the resource themselves.  Fish must be harvested with considerable effort in time, energy and personal risk often far away from where the fish will be consumed. The most efficient means of matching buyers and sellers is for them both to congregate in a location that maximizes the efficiency of all actors. Therefore it’s far more efficient for fish to converge on Billingsgate market in London than it is for the concentration of buyers in London to visit all the ports where fish are landed. The same dynamics are certainly not at play in the cloud computing.

Accessing cloud services requires little expenditure of energy on the part of the consumer — no more than a few clicks of the mouse. There is also very little effort required to find multiple suppliers of cloud services — Google “Cloud hosting services.”

As Zysman and I wrote in our recent position paper for the Transatlantic Policy Network that cloud is unique in being at once a competitive service, a dynamic enhanced utility, an ICT infrastructure/platform and innovation eco-system, a marketplace, and a production environment. The cloud is its own market.

Advocates for the creation of new primary markets for cloud services will argue that a lack of standardization in cloud computing resource definitions creates an opportunity for markets to arise. There is no benchmark today against which cloud services from different providers can be compared: There is no cloud equivalent of West Texas Intermediate crude. This makes it very difficult for consumers to price compare across cloud providers. This information asymmetry should create the opportunity for a market based on common benchmarks for cloud resources.

I completely agree that the lack of standardized — comparable — cloud resource benchmarks is a problem. The question then becomes how much value could a 3rd party market maker extract based on that information asymmetry. My view is not enough to be viable.

If benchmarking standards become a consumer requirement then cloud providers will adopt them, but the energy expenditure by consumers to compare costs is just not that great. Now I have to click three times rather than just once. Or more realistically volume consumers of cloud services will programmatically make three API calls to obtain the current benchmark spot price for the resource they need.

The opportunity for 3rd party market making in cloud resources is so small as to be essentially non-viable. As a consumer there’s just not going to be enough value in the service for me to accept the market maker’s vig on the transaction.

The internet has been the great disinter-mediator across many industries. The retail supply chain — with lots of actors between suppliers and consumers — has been completely disrupted by Amazon and their direct to consumer model. Intermediate actors in value networks create information asymmetries — each actor has access to unique information on the buy and sell side of the transaction and these asymmetries create sub-optimal exchange of value between the producer and the consumer.

The cloud business models of Amazon, Google, Microsoft and others are premised on having the shortest path between supply and demand. None of these companies is going to willingly permit the insertion of a 3rd party market maker and no rational consumer would want that either. A for profit 3rd party market ultimately reduces margins for producers or increases prices for consumers — or both. It cannot be any other way.

Secondary Markets

Secondary — derivative — markets emerge when there is uncertainty about either supply or demand i.e. there is imperfect information about the state of the primary market or information asymmetry between buyers and sellers. This uncertainty can be traded or arbitraged. Traders in derivative markets are betting that they have better information about future supply and/or demand than other actors in the market.

Futures contracts allow buyers and sellers to hedge risk driven by volatility in either supply or demand of a particular commodity. In essence the contract allows you to fix the price for future delivery of a commodity. The contract itself can then be traded based on whether other actors believe the actual market price — supply or demand — will be higher or lower than the contracted price. The more volatile a market is the greater the opportunity there is for arbitrage.

Wheat futures contracts have a straightforward economic rationale. Volatility is largely driven by the unpredictability of seasonal weather patterns. Supply — crop yields — can vary dramatically based on whether there is a particularly hard or long winter, a snap hard frost late in the spring etc. Both farmers and breakfast cereal manufacturers use futures contracts to lock in prices and hedge risk for their individual rationale. Derivates traders make a living by trading on the lack of perfect information on either future weather or consumer consumption patterns.

Trading in petroleum products adds additional factors which drive volatility. Crude oil — what comes out of the ground — is ‘cracked’ to make make many different products; petrol, naptha, diesel, heating oil and feedstocks for a wide arrange of industrial chemicals. The efficiency with which crude can be turned into these products depends on its ‘quality’ i.e. how heavy it is and the level of contaminants — sulfur etc.

Petroleum based futures contracts are based on standard quality benchmarks — West Texas Intermediate being but one. Out of the ground crude production is fairly predictable at a macro level — OPEC sets production quotas for their producers. The volatility which drives the rationale for futures contracts is based on variation around the quality benchmarks, seasonal demand for petrol, heating oil etc. based on weather patterns and economical cycles driving macro demands for industrial chemicals and lubricants etc.

The problem as I see it for anyone wishing to create a futures market in cloud resources is where’s the volatility?

There has to be some driver of volatility in either supply or demand for there to be an economic rationale for futures contracts. Unfortunately for nascent 3rd party market makers there is no indication at all that the market for cloud resource provision shows any meaningful volatility.

On the macro demand side of the equation there is little or no volatility. We’re at the start of the adoption of cloud computing and consumption demand shows only one trajectory: up. There is no volatility on the supply side either. The major providers of cloud services — Amazon, Google, Microsoft, Rackspace and others — all build far ahead of the demand curve. They also have almost perfect statistical insight into demand patterns driven by data from millions of direct consumer relationships. There is little or no information asymmetry in this value network.

Of course it would be correct to point out that individual cloud providers do see changes in demand over varying time horizons. The demand for cloud resources powering online commerces peaks during the christmas holiday season and tails off during the summer. But the cloud providers themselves already leverage these demand fluctuations by providing spot pricing for their resources.

Consumers can contract for the provision of a certain amount of compute capacity at a fixed price if they are willing to be flexible about how long it takes for the capacity to be made available. Its likely that all cloud providers will eventually offer the same. This example again demonstrates that the cloud is its own market and that there will be very little — if any — opportunity for 3rd parties to insert themselves between suppliers and consumers.

We have witnessed a major shift over recent decades towards a digital services based economy[1]. Exponential growth in the power of information technology enhanced by Internet driven network effects mean that even the most mundane manufactured products are seen as just one component of a broader digital services based value chain.

The shift to digital services has transformed the global competitive landscape. Maintaining competitive differentiation requires continuous innovation – and lifecycles are contracting. In other words change is a constant.

The dynamics of a hyper-competitive global market mean that static and long-lived models of company structure, operations and information technology services are no longer fit for purpose. We need a new operating model for the enterprise.

It is a reality of business today that supplier relationships, logistics networks, product and service design and customer service all survives in a state of permanent flux. Any path to sustainable competitive advantage will require a high degree of operational adaptability.

Business functions, processes, organizations, supplier relationships and technology need to be seen as building blocks that can be re–configured as needed to address changing competitive landscape.

This new Component Operating Model (COM) requires a ‘Lego Brick’ approach to designing and implementing processes and the organizations that support them. Implementing a COM based approach will have profound impacts on the structure of organizations, the nature of work.

Business designs based on COM will create significant stress for traditional IT infrastructures and organizations. Our current IT services were built to serve a static – and often functionally silo’d – operating model. IT needs to become much more dynamically adaptable to keep pace with the speed of business today.

A new Component Architecture Model (CAM) approach to IT infrastructure, applications and services will be required to ensure that IT can deliver what the business needs. The time between identifying a business need and delivering the required IT solution needs to becomes hours and days rather than months and years.

The ‘Ying and Yang’ of COM and CAM is a radical departure from the discontinuities of the past. We are about to enter an era where — for the first time — business and technology operating models can be seamlessly aligned to enable fast, flexible response to a rapidly changing competitive landscapes.

Combining the approaches of both COM and CAM will enable us to view the entire portfolio of business functions, processes, organizations, relationships and technology as a set of re-usable components that can be configured as demanded.

Welcome to the Composable Enterprise.

Radical Change

Most companies would consider the Composable Enterprise Model (CEM) radical today. And yet its difficult to imagine how more static approaches to business operations will survive in an increasingly dynamic and competitive global economy.

A ‘chasm’ already exists between those companies that grew-up-digital and those more traditional companies attempting to adapt to new digital markets.

Companies such as Amazon have always had to deal with hyper-competition and constant, fast-cycle change. Amazon’s approach to building out its retail business is driven by a deep understanding of digital business dynamics. Amazon is perhaps the closest proxy for a Composable Enterprise today and that makes it a formidable competitor for traditional ‘Bricks and Mortar’ retailers.

Building a Composable Enterprise operating model — where functions and processes can be continually re-configured with low cost and operational impact — raises significant challenges. The two greatest of these will undoubtedly be organization and technology. This is change management in extremis.

Change of this magnitude requires visionary and forceful top-down leadership. Implementing the Composable Enterprise approach is not going to happen from the bottom-up. The level of disruption in most organizations will require a vision to which staff can aspire. The greater the change the more compelling that vision will need to be.

Companies are unlikely to find that they have a large proportion of staff with the skills and capability to build, operate and thrive in the Composable Enterprise model. Significant re-skilling will be required.

Many already-digital businesses are likely to find that the demands placed on IT by the Composable Enterprise Model far exceed the capabilities of existing systems and application landscapes. This will be amplified for company’s attempting to transform from more traditional ‘analog’ business models.

The vast majority of IT systems in existence were designed for static business models and not for fast re-configuration of business processes, data flows and interconnection between disparate functions. As I will describe below the radical CEM based approach will require a radical technology fabric to match.

The Component Organization Model

Successful re-configuration of a company’s organization and processes to support the Composable Enterprise approach will require a radical departure from historic models of organization and process design.

The new Component Operating Model approach requires us to break processes and the organizations that manage them into their most fundamental building blocks. COM is based on the following key design principles that aim to maximize adaptability and the speed with which new processes can be configured and enhanced.

The Minimal Function – Compound business processes are broken down into the minimized set of ‘atomic’ functions. Each function can accomplish some aspect of the process independently of other functions and consists of the smallest organization possible to accomplish the function. The purpose of this principle is to ensure that individual functions can be improved or replaced while minimizing impact to the wider organization and business process.

Least Dependency – Individual business functions should be designed to minimize dependency on any other function. This principle ensures that failures in a single function do not bring the entire process to a halt. Process steps such as authorizations should be asynchronous to avoid creating serial dependencies.

Shared Knowledge – Knowledge (Data, information and context) is a resource shared across all functions. Data governance and security policies should be designed to maximize safe access to data to the broadest audience. Functions are prohibited from maintaining ‘Private’ data sets.  The holding and protection of ‘private’ knowledge is one of the key barriers to organizational agility. When knowledge is not openly shared then individual functions can create dependencies on their existence even when that is not optimal for the overall business.

Predictable Contracts – Each function defines the information and authorizations it requires, what work product will be output and provides quantified performance guarantees.  This approach aims to minimize the governance and supervisory overhead that are often layered into organizations to overcome a lack of predictability and performance between business functions.

Maximized Human Value – Human resource should be applied only to those tasks that cannot be automated: Front-line customer/partner engagement and service and where insight, analysis and qualitative decision making is required etc. This approach has two payoffs. It minimizes overall human resource expense and maximizes the value of roles in the business leading to higher job satisfaction and lower attrition.

We do not yet delve into appropriate governance approaches for COM but it is clear that ‘Kanban’ would be the most effective approach to overall process control in the Composed Enterprise operating model. A ‘Pull-through’ demand-forecasting approach should maximize the efficiency of the overall production process.

What is truly radical about CEM is – finally – for there to be an intimate and fully aligned relationship between business and IT. This has never existed in the history of enterprise computing. From the days of IT staff in white lab coats managing huge mainframe computers to the present highly distributed and virtualized application environments – IT has never been able to keep pace with the demands of the business.

The Component Architecture Model

Technology provides the ‘fabric’ – infrastructure, applications and services — that underpin CEM. Information is the lifeblood of any value chain in today’s global economy and that ‘blood supply’ should connect the farthest reaches of company’s global business processes.

The dynamic and configurable nature of the Composable Enterprise operating model places significant demands on the underlying technology fabric. This is likely to mean a radical departure from the legacy IT infrastructures and applications portfolios found inside most organizations today.

The majority of legacy IT environments were built to support a much more static business model. It was assumed that business processes would remain unchanged for long periods or at least be subject to only minor modifications. Applications were generally designed to support the needs of functional silos within the business: Manufacturing, Finance, Sales, Supply-chain, Marketing, and Human-Resources etc. The growing need to build horizontally integrated processes between these functions has significantly challenged IT organizations and budgets over the last decade.

The Composable Enterprise business model takes horizontal process integration to an entirely new level by assuming that operating functions; processes, products and services will be subject to continuous change and re-configuration.

Enabling the Composable Operating Model will require a radically new approach to enterprise IT. In effect the COM will require a matching Composable IT Architecture. The key attributes and capabilities for such Composable Architecture Model (CAM) would be:

Elastic Infrastructure – The ability to scale up and down IT infrastructure resources based on changing demand patterns.

Component Service based Applications – Implementation of business applications as a ‘Composition’ of finer grained ‘atomic’ services to enable rapid re-configuration and enhancement. This capability when combined with Elastic Infrastructure enables business applications and services to be scaled rapidly to meet the changing consumer or business demands.

Auto Everything – The technology fabric should be architected to require the minimum of human intervention. Dynamic re-configuration of services, elastic resource provisioning and operations should be fully automated.

Unified Master Data Model – Implementation of a business defined and governed master data model providing a single-source-of data truth across the business.

Integrated Analytics – All data to be used by the business for decision-making, consumer and market insights, process and operational analytics should be aggregated in one integrated repository. This ensures a single-source-of-truth for all data based business decisions.

People Services – The fabric should expose core services that enable role-based user authentication and user profile management to facilitate expertise discovery and rapid organizational re-configuration.

Process Services – The fabric should expose core services that facilitate process monitoring and workflow integration to facilitate rapid process re-configuration.

Data Services — All master and analytics data should be accessible through standards based data access services and should be consumable in both human and machine accessible form.

Any Time, Anywhere, Any Device Access – Operational applications should be made available using a web based ‘Software as a Service’ (SaaS) delivery model accessible from any device capable of running a modern web browser.

The purpose of these requirements is to ensure that IT applications and services can be managed and delivered with the same level of granularity as the business functions that use them.

The Enterprise as a Service

The CEM approach is based on the need to ensure seamless alignment between business operations and IT services because no matter how mundane – all business is digital today.

Historic disconnects between IT and the business — from a function and rate of change perspective — are addressed under the CEM approach by creating structural alignment between business operations and the technology architecture.

The design principles behind COM and CAM are identical: Decompose everything to the most ‘atomic’ level, enable independent execution, interoperate through clearly defined interfaces and share information openly. It does not matter whether you are designing a business function or software component the same principles apply.

The purpose behind these high level principles is straight forward – even if implementation may not be: To facilitate, rapid adaptability, response to changing demand patterns and flexible and robust operations.

Implementation of CEM requires the implementation of both COM and CAM. Combining COM and CAM enables us to view the Enterprise as a Service: A matched portfolio of functional, organizational and technology components that enable rapid re-configuration in response to future demands of a modern digital services economy.

The economic evolution of computing platforms appears to be guided by a number of ‘Laws’ that are independent of any specific underlying technology. These ‘Laws’ help explain the evolutionary trajectory of mature platforms and can act as a predictor for the trajectory of emerging platforms and behaviors of those that control them.

The first three laws of platform economics are: Value always migrates up the stack, Value gets integrated over time, Those that control platform evolution get to define how value is extracted. A definition of each of the ‘Laws’ is provided below:

First Law - Value always migrates up the stack — Value creating functionality which starts off embedded in silicon or hardware functionality will eventually migrate to software. As a platform matures value will be driven by software functionality that is increasingly abstracted from any specific underlying physical implementation. The canonical example of this law is the networking stack: Highly specialized network routing silicon gives way to specialized router operating systems which eventually give way to software defined networking layers in cloud computing fabrics.

Second Law - Value gets integrated over time - Functionality which adds value to computing platforms will — if it initially sits outside the control of those that develop the platform — be integrated into the platform. The greater the economic value the more likely that external functions will get absorbed and integrated. In the early days of Microsoft’s Windows platform many valuable functions were provided by 3rd party software — TCP/IP networking protocols, memory managers and anti-virus software to name three examples. Over time these functions were integrated to further enhance the value of the platform.

Third Law - Those that control platform evolution get to define how value is extracted - In other words the owner/developer of a platform will always act in a way that enhances the platform’s primary driver of economic value. Google will always make platform decisions that enhance its digital advertising based revenue model. The functionality of iOS will evolve to enhance Apples ability sell hardware and to extract value from consumers use of services and consumption of media. Microsoft chooses to implement ‘Do Not Track’ by default in IE 10 to both undermine Google’s advertising based business model and to enhance the attractiveness of Microsoft’s new device and services business model. Red Hat is the only open source platform company generating more than $1Billion in revenue per year.

Over the last couple of years the term cloud computing has become become public relations shorthand for “Look, look, we’re also cool.” The expression is now so overused and is applied so widely that it has become almost meaningless. That’s a problem. Senior business decision makers have been set a drift in a sea of PR and marketing confusion just at a time when they need to be making strategic technology choices that may directly impact the future of their firms. I offer this brief guide to cloud computing in an effort to help cut through the fog currently enveloping this important topic.

Let’s start with a formal definition of cloud:

The provision of computing infrastructure, platform or application service as a utility, which can be consumed by any Internet connected device, using open standard protocols where variability in demand is satisfied through the dynamic and automatic provisioning of pooled hardware, network, and software service resources providing the illusion of infinite scalability and are generally billed for on a pay-as-you-go basis.

Which can be consumed by any internet connected device, using open standard protocols — In general cloud based applications and services can be accessed by any device that can connect to the internet using standard internet protocols. Generally a web browser would be used by an end-user to access services. However, native applications on smart phones may also be the interface through which users connect.

Where variability in demand is satisfied through the dynamic and automatic provisioning of pooled hardware, network, and software service resource — This is one of the key differentiators of cloud computing which separates marketing spin from reality. For a service to truly be considered as cloud computing it must be able to automatically scale up and down to meet variations in demand. Hosting a website on an internet server — or even a group of servers — does not qualify as cloud computing because if demand exceeds the capacity of the underlying hardware then human intervention is required to either add additional servers or to move the workload to a larger capacity machine. Cloud services make these adjustments automatically.

Providing the illusion of infinite scalability - Users of true cloud computing services never experience a lack of resource due to an excess of demand. Cloud provider organizations constantly build out their infrastructures so that available capacity always exceeds near-term user demand. It is the ‘Illusion’ of infinite capacity that separates cloud architecture from more traditional forms of hosted computing infrastructure.

Generally billed for on a pay-as-you-go basis — Though not universally the case, the charge for cloud computing services is normally based on the amount of computing resource consumed over time. This is the model general used for IaaS services. This compares the fixed or contract pricing models used for more traditional hosted application services.

True cloud services should meet all the the criteria outlined above but in particular they should be able to provision new services and scale automatically based on fluctuating consumer demand without human intervention. Services being sold as “Cloud” which come with fixed capacity limits or obligations to purchase defined hardware configurations are — by definition — nothing of the sort.

There is one other aspect to cloud computing that creates significant confusion: How it is deployed? There are three accepted models of cloud deployment: Public, Private and Hybrid.

Public Cloud is where an application is deployed on a cloud infrastructure that is shared by multiple customers and is directly accessible over the internet. The most commonly used public cloud service is Amazon’s Web Services. Other public cloud environments include Microsoft’s Azure and Google’s App Engine. Unfortunately the Public deployment model is often made synonymous with the term cloud computing. It is not.

Private Cloud is a cloud infrastructure that is dedicated to one particular customer. The environment may be run behind a firewall in the customer’s own data center or it may be run for the customer by a third party. In either case the private cloud resources are dedicated to the needs of one company or organization. Due to privacy and security concerns, government’s are likely to create private cloud infrastructures that are shared by multiple agencies or ministries.

Hybrid Cloud is a deployment choice where public and private cloud infrastructures are linked together. A company may choose to run certain sensitive workloads on its dedicated private cloud infrastructure while running less sensitive workloads on a public cloud infrastructure. The two environments are joined together to give application users a seamless experience. A company may also temporarily utilize public cloud infrastructure capacity to help manage peak demand that exceeds the capacity of the company’s private cloud infrastructure.

We now have a complete definition of cloud computing: Cloud Computing is a specific form of systems architecture — meeting the criteria outlined above — plus a particular approach to resource management and application development and delivery. This cloud architecture can then be deployed either publicly on a shared infrastructure, privately on a dedicated infrastructure or in a hybrid of public and private.

Diffusing the Cloud: Cloud Computing and Implications for Public Policy takes a deeper dive into the world of cloud computing examines it’s evolution and some of the economic and policy issues surrounding this important new domain.

Update: Since writing this last night @fscavo has written an excellent blog on the fallacy of Oracle’s cloud offerings. Well worth the read!

Part 2 of this series of articles looked at how transitioning from scarcity to an abundance of fundamental computing resources enabled the historic one-to-one relationship between operating systems, applications and underlying hardware to be broken. Part 3 will examine how the ability to decouple hardware and software evolved into a new strategy for managing IT systems — saving company’s millions of dollars in the process — and laid the foundation for today’s cloud computing architectures.

The Mac on which this article was written can run not only Apple’s native OS X operating system, but also Microsoft Windows, multiple variants of Linux and — if needed Oracle’s Open Solaris operating system. A modern personal laptop computer has such an abundance of computational capacity that multiple distinct hardware architectures can be emulated in software.

This ability to virtualize the link between hardware and software with little or no performance penalty is a direct benefit of the economics of abundance. New layers of virtualization software allow us to run a copy of Microsoft Windows or Linux on a Mac computer alongside the Mac’s native operating system with each system sharing the same computational, storage, memory and network resources. This ability to decouple hardware and software provides significant economic benefits.

The ability to run multiple — but distinct — software architectures on the same physical hardware allows companies to amortize hardware costs across a wider range of application workloads. Virtualization is interesting and helpful on a laptop, but is truly transformative when applied to multi-million dollar computer server infrastructures.

Under the economics of scarcity a single enterprise software application would typically be tightly coupled to a single hardware platform. Often the underlying hardware would be running at only a small fraction of its available capacity. The ability to run multiple application workloads on the same hardware through virtualization enables IT departments to manage their hardware resources much more efficiently by running them at higher levels of utilization.

The use of virtualization strategies is now standard practice in many — if not most — enterprise computing environments. In many cases a few hundred servers running virtualized workloads have replaced 1,000s of servers previously dedicated to individual workloads. Applying this approach has saved many millions of dollars across enterprise IT departments.

At least one company — VMWare — became successful by meeting the market demand for virtualization solutions and was perhaps the first child of this new era of resource abundance. VMWare is no-longer the only player. The economic benefits of this approach to software architecture have spawned a diverse and  increasingly competitive ecosystem including offerings from Microsoft, Oracle, Citrix and Parallels and the open source based Xen.

The ability  to turn the compute, program memory, storage and network resources of many independent servers into a virtualized pool of resources to be allocated as needed by software applications has laid the foundation for emergence of Cloud Computing.

Part 4 of this series will look at how the commoditization of computing resources and the ability for them to be managed and allocated on demand forms the foundation of new cloud computing architectures and a radical change in the way software applications and services are created, managed and consumed.

Part I in this series of articles outlined the impact that the economics of scarcity has had on both software architectures and the structure of the computer industry over the last forty years. Part II of the article will discuss the transition from the economics of scarcity to the economics of abundance and how profoundly that has altered — and continues to alter — the computing landscape.

The introduction of the IBM PC in the early 80’s laid the foundation for a transformation in the economic model of the computing industry. IBM broke with its traditional model when it outsourced the supply of the computer processor and operating system for its Personal Computer (PC.) Little did IBM know that this single decision would transform an industry. The decision to source processors from Intel and an operating system from Microsoft and to not bind both these suppliers to exclusive supply agreements created the opportunity to build a 3rd party PC industry based IBM’s generic blueprint. The market for PCs grew and other players such as Compaq created competitive alternatives to IBM. The highly competitive dynamics of the PC industry based on the open hardware blueprint created by IBM had very direct consequences.

Rampant competition within the nascent PC industry kept prices under control and yet each vendor’s hardware ran the same DOS — and subsequently Windows — operating system licensed from Microsoft. This in turn meant that every PC sold became an available ‘socket’ for a growing ecosystem of application software developers. The same application would run without modification on every model of PC. The combination of these factors lead to exponential growth in the sales of PCs and this growth lead to vast economies of scale in the production of microprocessors, disk storage and memory chips. All of this was enabled by dramatic improvements in hardware fabrication technology – largely following Gordon Moore’s law. As a result we have witnessed an approximate doubling in computational capacity, program memory and disk storage capacity every 12–18 months continuously for over forty years.

In the early 2000’s the computing industry reached a tipping point where hardware capacities started to outstrip the demands of application software. The industry was about to transition from economics of scarcity to the economics of abundance. The impacts of this transition were — and continue to be — profound. The economics of abundance mean that hardware resources can be viewed as essentially unlimited. It is interesting to note that it was at this time that Intel entered the software business. It did so because it felt the need to incubate new software technologies which would utilize the full capacity of the chips it was producing. If processor capacity far outpaces the demands of Microsoft Windows, Office and a Web Browser then consumers do not need to upgrade their machines that often — and that is bad for Intel’s business.

It is hard to understate the impact of this transition from scarce to abundant computing resources. The iPhone is a child of this transition. The economics of producing a completely integrated, highly functional, software services based smartphone could only be achieved in an era of hardware resource abundance. An environment where underlying hardware resources are – to all intents – a limitless commodity changes the calculus of software development. It is helpful to consider the Linux operating systems as an exemplar of these changes.

During the period when the economics of scarcity reined the most direct analog to Linux was its precursor the UNIX operating system invented at AT&T’s Bell Labs. UNIX was designed to be a general-purpose operating system which could run on many different types of underlying hardware. However, when hardware resources are scare the operating system needs to be customized to be as efficient as possible on each different hardware platform. But tuning UNIX in this way created so much divergence across the UNIX family that it could no longer be considered a single operating system. Applications written to run on one variant of UNIX were not guaranteed to run on any other variant. The UNIX ecosystem never developed or benefited from the same economies of scale witnessed by the IBM PC ecosystem.

Linux on the other hand is an operating system which benefits from the economics of abundance. So much computational capacity is now available that some of this capacity can be “wasted” by layers of software designed to decouple application developers and users from differences in underlying operating system and hardware platforms. This ability to “waste” computing cycles on decoupling means that Linux truly has become a general purpose operating system. From a historic perspective Linux’s hardware adaptability is truly astounding. You can now run a single operating system on almost any hardware architecture available; from the smallest Raspberry Pi to the largest IBM Mainframe machine. In general an application written for one Linux distribution will run unchanged on any other. Such adaptability would have been unthinkable in the era of scarce hardware resources.

This Linux examples still relates to an environment where there is a one-to-one relationship between a computer – hardware – and the operating system and application software running on it. Applications written for the Linux operating system still only run on Linux and will not run — without major modification — on Microsoft’s Windows or Apple’s OS X operating system. But computing resource abundance has now enabled even this problem to be solved, allowing multiple different operating system and application ecosystems to happily share a common underlying hardware platform.

Part III of this article will examine how the ability to now share abundant hardware resources across many application workloads is transforming computing architectures and ultimately lead to the emergence of cloud computing.

The underlying economics of computing resources have always had a profound impact on development of computing architecture and in-turn the structure of the computer industry. In this regard the emergence of cloud computing is no different. Cloud computing has emerges as the product of a fundamental transition in the underlying economics of computing resources and — as in the past — this economic transition will drive profound changes in the structure of the computing industry. The nature of this change can best be described as a transition from the economics of scarcity to the economics of abundance.

Until relatively recently all fundamental computing resources were considered scare. Processors had limited computational capacity and were constantly being outstripped by the demands of software. Limited and expensive computer program memory and disk storage put severe constraints on the size and complexity of computer applications. The earliest network connections could only transmit data slowly and at very high cost. The cost and limited capacity of these foundational components of computing infrastructure – computation, memory, storage and network bandwidth – placed severe limits on the complexity of operating systems and the layers of software which ran on top of them.

In a world of highly constrained hardware resources software needs written for specific hardware architectures and configurations. In the past this tight coupling between software and hardware was essential to extract every last ounce of performance from very expensive systems. IBM produced highly optimized operating systems dedicated to specific mainframe hardware, memory and storage architectures. In his seminal book The Soul of the New Machine, Tracy Kidder describes the very intimate relationship between the development of hardware and software for one of the earliest DataGeneral mini-computers. Digital Equipment Corporation did the same for its each generation of its PDP and VAX mini-computers. Even later generation companies such as Sun Microsystems followed the same approach with its Solaris operating system to extract every last ounce of performance from their workstation and server products.

The early success of Sun Microsystems in markets for engineering workstations and high performance servers for the financial services industry show just how successful this “Tight-coupling” of software and hardware was as an economic model in the computer industry. While the demands of software developers and their algorithms outstripped the capacity of hardware resources this tightly coupled model would hold sway. For the last forty years the economics of scarcity have defined both platform architecture and the very structure of the industry itself.

Customers have always been willing to pay a premium for software and hardware solutions that could optimize scare resources and keep operating costs to a minimum. Oracle, Cisco and EMC are but three companies whose fundamental value proposition was based on the economics of scarcity. Each of these companies created value by being best in class in helping customers optimize and manage network bandwidth, database performance and scale and disk storage capacity respectively. They could each charge a premium price for their products and services because the costs of the underlying infrastructure resources were so high that marginal improvements in resource optimization lead to significantly positive outcomes in cost control.

The economics of scarcity have had a profound effects on evolution of computing architecture. Operating systems were tuned for specific hardware platforms. Databases were tuned for specific operating systems environments. Application performance was optimized by targeting specific database technologies. It is interesting to footnote that the fact that we have historically placed application code inside databases — in the form of ‘Stored Procedures’ — is a direct result of the need to optimize for scare resources rather than as an elegant architectural design decision. Finally, all of this underlying application architecture was then accessed by tightly coupled client interfaces using highly optimized network access protocols. This model of the fully — vertically — integrated application architecture dominates today’s enterprise computing landscape and does so as a direct result of the economics of scarcity.

The vertically integrated application model may have helped optimize for very scarce and expensive hardware resources but it also lead to significant future costs through fragility and lack of both flexibility and interoperability. Vertically integrated application architectures are very rigid structures. The tight-coupling between layers makes it extremely difficult, costly and often impossible to replace components of the architecture with competing offerings. Migrating an SAP instance originally implemented on Oracle over to Microsoft’s competing SQL Server product is a non-trivial and expensive proposition. Swapping out IBM’s WebSphere middleware product for a competitor such as JBOSS or Weblogic is a task only the most fearless of IT organizations would take on.

The built in switching costs of such highly optimized application architectures had the effect of cementing the early success of today’s leading enterprise software players. These switching costs acted as an accelerant to local network effects within individual vendor ecosystems. IT departments are typically organized around these ecosystems i.e. the IT organizational model becomes vertically integrated as a response to the application architectures they support: The Oracle center of excellence only supports Oracle based applications. The Microsoft center of excellence only Microsoft platforms etc. Unfortunately most organizations are dealing with the ugly reality of the cost structures and lack of flexibility that this approach creates. An approach to optimizing scare computing resources has ended up creating higher acquisition costs, very high switching costs, much higher systems integration costs, dramatically reduced application adaptability and centralization of industry power in the hands of a few select vendors.

Part II of this article will look at how a transition from the economics of scarcity to the economics of abundance in the early 2000’s lead to a transformation in platform architectures and lead the foundation for the emergence of cloud computing.

Imagine a world where all internet browsers are required to present users with a simple question: “Do you want your online browsing activity tracked, recorded and shared with marketing companies — Yes or No?”

What percentage of users do you suppose would answer “No?” My guess would be greater than 90%. If I’m anywhere close to being correct then Microsoft’s controversial decision to enable ‘Do Not Track’ (DNT) by default in IE10 and Windows 8 would seem to be very much inline with consumer sentiment. But advertisers, Yahoo, and the developers of both the Apache web server and FireFox browsers are all decrying Microsoft’s decision.

This relatively arcane debate over a new internet standard masks a much more critical issue: In the long-term, how viable is the internet’s — advertising based — ‘Free’ content model?

Advertising revenue is the internet’s predominant business model. This is of course the core business model that enabled Google to become the 800 lb. gorilla of the industry. But this business model is based on users accepting a ‘Faustian’ pact where they agree to increasingly invasive tracking of their online activities in exchange for free content.

There’s only one small problem with the current situation. Very few consumers have ever read or consented to the ‘Contract.’ Most of consumers have no earthly idea how invasive today’s internet tracking technology is and once they are aware they are not going to like it.

The howls of outrage from advertisers and from sites like Yahoo — that depend on advertising revenue for their very existence — about Microsoft’s decision to enable DNT by default mask a much deeper fear. Once online consumers realize the true value of their ‘click-stream’ perhaps they will start asking for a lot more in return. Maybe the fallacy of ‘free’ content on the internet will be made transparent with all the dire consequences that would entail for the companies that depend on this business model.

Here’s a suggestion. If advertisers — and others — are so concerned that DNT should require an explicit opt-in by consumers then lets implement the solution I alluded to in my open question. Every browser should prompt the user to make an explicit choice for the DNT setting during installation or upgrade. How could that possibly be a bad thing? It would avoid taking Microsoft’s assumed default approach and yet it would ensure that every consumer was made aware of the option and given a choice.

I suspect that advertisers would absolutely hate the idea of requiring an explicit and visible opt-in for the simple reason that they already know that the most users would opt not to be tracked.

All of this raises profound questions about the long-term viability of the internet’s only proven business model. I suspect that the implementation and awareness of DNT is going to force many consumers to confront their ‘Deal with the devil’ and many are not going to like the bargain they have struck.

As consumer resentment over invasive click-stream tracking grows the viability of purely advertising based content business models will come into serious question.

Companies that depend solely — or largely — on online advertising driven revenue streams — Google, Yahoo etc. — have a major strategic problem. How will they survive if the majority of consumers — as I suspect — decide they do not want to be tracked? The undermining of the current model of free content in exchange for your behavioral data will require a re-think of internet business models.

Consumer awareness about the value of their behavioral data — highlighted by options such as DNT — is likely to make paid content models much more attractive. Smart paid content providers will find ways to amplify consumers fears about being tracked. Like the emperors new clothes ‘Free Content’ will be seen to be not free at all. Consumers will then need to decide whether a much more transparent contract — where they pay to access content without giving up their privacy — isn’t an entirely more acceptable way to browse.

Ps. I do find it ironic that the Mozilla appear to be siding so enthusiastically with the position of advertisers in the DNT debate. Might that have something to do with the fact that 98% of their $121Million in 2011 revenue came from advertising based search royalties?

Pps. Here’s a prediction. If the internet/advertising industry does not voluntarily implement the proactive DNT choice notification I suggested above then they will find themselves forced to do so within the next five years. I would not be surprised to learn that European Commission bureaucrats are just waiting for the DNT specification to be completed and ratified before making this demand of all browser manufacturers.

We are currently witnessing a major pivot in Microsoft’s core business model. It is starting to become clear that — as Steve Ballmer recently announced — Microsoft is deadly serious about becoming a global leader in consumer “Devices and Services.” Successful execution of this strategy will require the company to control everything from manufacturing, distribution logistics through to retail.

The company appears to be focused on executing a ‘Leader’ strategy in the devices business which would give them even greater control than Apple famously does over the end user experience. In addition the company is likely to repurpose its online service investments to add value to this new device-led strategy.  A-la Apple, consumers will have no choice but to use Microsoft own services when using a Microsoft mobile device and competition authorities will be powerless to prevent it.

If correct, this pivot will has profound implications for the structure of the company, shareholder value and for the entire mobile technology industry.

In the beginning…

I was reminded recently by a couple of former Microsoft colleagues about a presentation Steve Ballmer gave to the sales force several years ago. It was Steve’s keynote at the annual sales meeting and it was one of the more emotional presentations by Steve that I or any of the other attendees had witnessed. The theme of Steve’s presentation was Muhammad Ali’s legendary ‘Rumble in the Jungle’ with George Foreman. In his presentation Steve — a huge sports fan — analogized Microsoft to Ali — an underdog in the fight of its life against younger, fitter and more agile competition. Ali of course won the fight supported by 60,000 fans chanting “Ali, bomaye! Ali, bomaye!”, “Ali Kill Him.” You can picture Steve energetically bounding around the stage screaming “Microsoft Bomaye!” getting 15,00 Microsoft sales staff wound up into an enthusiastic frenzy.

On face value this revving up of the sales team was vintage Ballmer, ever the master salesman himself, getting the troops pumped up to go do battle with the enemy. But recent conversation have caused me to suspect that there was a much more important hidden message. Ali did not enter the ring in Zaire to fight Foreman without a plan. Far from it. In fact Ali was very clear before the fight that his strategy would be to absorb all of Foreman’s punishment in the early rounds, tire him out and then land a killer knockout blow. That’s exactly what he did.

I now suspect that Steve came to that annual sales meeting with a long-term strategic plan already developed. Steve was pumped up because he realized that Microsoft would need to live through a punishing period where everyone would write-off the company, competitors would land punishing blows and eventually — when the time was right — Microsoft would — like Ali — deliver a knockout blow.

After years of punishment by the likes of Google and Apple we may now be approaching the point where — at least in Steve’s view — Microsoft is about to land a killer punch.

Out of Advertising

At another annual sales meeting a few years after “Microsoft Boomaye!” my team met with Steve to discuss a number of technology policy challenges we were facing. We felt there was an opportunity for Microsoft to take the high-ground on personal data privacy issues as a competitive position against Google. Steve agreed that this was clearly an opportunity but while sympathetic to taking that position he was not yet ready to support it. He felt that putting a stake in the ground at this stage would be premature. Once the company committed to a policy position it needed to stick with that position. Steve felt that putting a stake in the ground on consumer privacy before the online advertising market had matured might unnecessarily limit the company’s future options.

Three recent announcements clearly indicate that Microsoft is now ready to put that stake in the ground and that the company’s experiment with an independently targeted advertising based revenue model is over.

The company’s write-off of the $6,2 Billion investment in aQuantive was the first indicator.  Microsoft made than investment as a competitive bulwark against Google and now its being written off. In effect the online advertising business — at least in its current form — has been judged to not be core to Microsoft long-term strategic success. This not to say that the company will withdraw from the advertising business altogether but I believe Steve and Co. have a different and more focused model in mind.

Microsoft’s May 2012 announcement that the next version of Internet Explorer would enable “Do Not Track” by default has caused outrage across the advertiser community. That is certainly one very large and very clearly marked stake in the ground. In the most direct way it can Microsoft is saying to advertisers that they are no-longer perceived as important to the company’s long-term success.

Finally, The company’s announcement this Monday (October 22nd 2012) — clarifying that it will not use consumers personal data for targeted advertising- is a final nail in the coffin for the once imagined head-to-head battle with Google for advertiser’s dollars. That game is over and Microsoft is moving on with what it believes is a far more compelling advertising augmented revenue strategy.

Into Devices

Steve Ballmer’s recent announcement that Microsoft should now be considered a “Devices and Software” company was met with derision by the digerati and commentariat. The company’s track record in devices — beyond keyboards, mice and the XBox — is lamentable. Zune players anyone? Even XBox — a huge success by any criteria — had its share of manufacturing problems which have not exactly burnished Microsoft’s hardware credentials. But what if everything to date was — as they say — a learning opportunity?

It famously takes three version for any Microsoft product to reach its full potential. Its quite possible that version three of “Microsoft as a device manufacturer” is about to be unleashed on the world. The evidence for this can be pieced together from a number of moves over recent years. Apple’s acquisition of Intrinsity — an ARM processor design company — in 2010 generated a huge amount of commentary but — strategically — not such a surprising move by a hardware company. What might be more surprising is knowing that Microsoft established a Silicon Architectures  “SiArch” team long before Apple got into that business and has been hiring some top chip design talent.

In July 2010 Microsoft signed a new “Architecture” license with ARM. This license gives Microsoft the rights to design its own processors based on the ARM architecture. Just the sort of thing you would need if — like apple — you want to be in the business of creating differentiated processor capability for your own range of devices.

This week saw the publication of a story claiming that Microsoft was manufacturing the new surface tablets in its own manufacturing facility in China. The tenuous nature of the sourcing for this story lead many to question its validity. I’m surprised by that reaction because Microsoft confirmed that they has dedicated manufacturing capabilities back in 2007.

From what I can piece together I suspect the company’s manufacturing capability has advanced significantly over the last five years. It would not surprise me to learn that Microsoft has invested in the level of massive dedicated manufacturing capability that would enable it to compete as a global device manufacturer. Anything less than that would be a ‘follower’ strategy and I don’t think the Microsoft leadership team see themselves as followers.

I believe that the pivot towards a “Devices and Services” business strategy is premised on the company becoming a category leader, not a follower. If that is the case then they would want to have much more control over supply chain and distribution logistics than even Apple achieves. Instead of merely assembling components a truly agressive ‘Leader’ strategy would have the company moving upstream into component production. We already know that the company has the ability to create custom processor designs. Are those really nice surface tablet displays also designed and perhaps even manufactured in house?

Becoming A Consumer Business

I have been a long and constant critic of the Microsoft management team for their lack of focus and obsession with becoming a consumer company. I would not be the first to suggest that — over the last 10 years — Microsoft has become the ‘Rodney Dangerfield’ of the technology industry: Unable to get respect from anyone.

The reality is that by revenue and profitability measures alone Microsoft is quite clearly an enterprise software company — not a consumer products company. The evidence presented indicates that the company maybe getting ready to finally bring clarity to its schizoid business model with a massive pivot towards a “Devices and Service” consumer facing business model. That pivot would require very significant investments in an end-to-end value chain.

We know that Microsoft has device design and manufacturing capability. We also know that the company has invested in its own retail presence over the last few years. That was viewed as an attempt to mimic Apple. But what if this retail strategy had always been about creating a storefront for the company’s own manufactured devices. Such a strategy would allow Microsoft to by-pass traditional low-rent — but margin eating — outlets such as Best Buy.

If the company is focused on becoming an end-to-end device manufacturer then the retail strategy takes on a much larger significance. With in-house manufacturing capability at one and and retail distribution at the other Microsoft would need to join the two. A ‘Leader’ strategy would require the company to take far greater control of distribution logistics. A ‘Leader’ would not sub-contract logistics to a 3rd party but rather would build an end-to-end managed distribution network from manufacture to consumer hands.

If Microsoft is committed to taking control of its own distribution network then hiring of Kevin Turner — a former senior WalMart executive — makes a lot more sense. Turner was viewed as a ‘square peg in a round hole’ overseeing Microsoft’s — core — enterprise sales business. During my time with the company Turner was widely disliked by Microsoft’s enterprise sales staff and was viewed as ‘Not getting it’ or understanding what was required to meet enterprise customer’s expectations.

What Turner does understand — from his years at WalMart — is consumer retail and supply chain management. These are just the skills to have on hand if Microsoft’s is about to pivot it’s business model towards a consumer “Devices and Services” company.

Serving The Device

Microsoft’s moves to put a stake in the ground regarding online consumer data protection change the company’s calculus about advertising as an intrinsic revenue stream.

The company clearly sees an opportunity to put Google on the back-foot. Microsoft can position itself a champion of consumer data protection and rights while placing Google in the uncomfortable position of either defending its agressive monetization of consumer data or backing away from its current business strategy. Neither of those are palatable outcomes for Google and the latter is certainly untenable given how dependent Google is on it’s advertising revenue stream.

To say that Microsoft has not performed in online consumer services would be a gross understatement. Bilions and billions and billions of dollars have been invested over many years with the company gaining almost zero traction in any meaningful service category. The investment in Bing search and the partnership with Yahoo have hardly yielded the type of market redefining results both company’s had hoped for. If the company is in the midst of pivoting its business model towards a consumer devices-led strategy then expect the company’s online services to be re-focused and re-purposed to that end.

Under this new device-led strategy online services would become value enhancers for devices rather than stand-alone businesses in their own right. Today services like Bing have to compete in an open market where consumers have choice. When a consumer is using a Surface device then the only search, maps or music service they will have access to will be those provided by Microsoft.

Microsoft has had its fair share of anti-trust problems with both the United States Department of Justice and with The European Commissions Competition authorities. In each case Microsoft was judged to have mis-used its monopoly position in the market for PC operating systems. No such claims will be possible against the company when it enters the device business for two reasons. Firstly, Apple and Google already dominate the market for mobile devices. Microsoft quite clearly would not have a monopoly to abuse. In addition Microsoft has rather cleverly cleaved its operating system strategy.

Win8-RT is an entirely different operating system from classic Windows. If Microsoft ties its own services to Win8-RT then competiton authorities are going to have a hard time make any abuse of monopoly case stick.

Wither the OEMs

In a recent article I concluded that Microsoft’s ambitions to be an true consumer products company would remain an expensive fantasy until Steve and the management team were willing to “Throw the OEM part­ner com­mu­nity under the bus.” It now appears that the company is ready to do just that.

Microsoft’s public pronouncements about the importance of the OEM in the wake of the release of Surface are more than a little disingenuous. Microsoft finally does appear to have crossed the rubicon in regards to the long-term value of their OEM partners. The company appears to embarking on a strategy that — while accommodating the OEM channel — is not going to throw them a life raft. OEMs who do not already have an alternative business strategy when Microsoft becomes their largest and most agressive competitor are in deep trouble.

Ironically the OEM channel is about to commit collective suicide by enabling Microsoft to become their executioner.

Microsoft desperately needs the OEMs right now. They are needed because that’s the only way to get Windows 8 into the hands of hundreds of millions of consumers. But — wittingly or unwittingly — the OEMs are acting as the vector for the “Virus” that will ultimately kill them off. Once a critical mass of consumers have become hooked on Windows 8 Microsoft will have established the reach and credibility of the operating system that is required to execute the company’s devices strategy.

Once Microsoft has its own volume device capability it will have no strategic need for the OEM channel. Microsoft’s message to the OEMs can be summed up as “Enjoy the ride while it lasts.”

Nok Nok…

Perhaps the company most likely to be impacted by Microsoft consumer devices strategy is Nokia. I have absolutely no doubt that Microsoft is going to become a manufacturer of a full range of Windows 8 based smartphones. That is going to happen sooner rather than later. When Microsoft launches a ‘Surface Phone’ it is quite simply game over for Nokia.

Consumers already have a hard time choosing betweens phones running Windows Phone 7 and competing offers from Apple and the Android OEMs. If Microsoft comes to market with a range of world-class devices running Windows 8 and sells them through its own retail channel then I don’t see much opportunity for Nokia  - or any other Win8 phone OEM for that matter.

Nokia’s strategic options are shrinking rapidly. It’s my view that when Microsoft enters the phone business with its own devices Nokia will have run out of time as an independent company. The most likely final chapter in its storied history sees Microsoft buying up Nokia’s sizable patent portfolio, its in-house design and engineering expertise and — perhaps — the brand rights. Perhaps Nokia’s CEO Stephen Elop will even rejoin the Microsoft leadership team to run the phone business. Alternatives to this outcome seem to be very few and far between.

Truly a sad end to a once dominant company.

Two OS to Rule Them All?

Can Microsoft maintain its current corporate structure if it successfully pivots its business model to become the leading devices and services company? I suspect not.

Successfully executing on the strategy outlined in this article  would clearly give Microsoft a new lease on life. On growth prospects alone you would want to own the stock. However, effecting this level of radical change in strategy would amplify — rather than mitigate - the schizoid nature of the company’s operating model. Microsoft would become a company of two polar opposites: A defined, successful and moderate growth enterprise software company at one extreme and a very high growth devices-led consumer business at the other. A single stock ticker representing two such radically different business models makes little sense.

One of the many strategic issues posed when considering splitting up Microsoft has always been ‘What to do with the Windows business.’ It’s a relatively trivial exercise to bucket Microsoft’s business divisions into either the enterprise or consumer category. But the Windows division quite clearly serves both.If you were to split the company up then where would the Windows division end up?

The problem of where to put the Windows division would be made a lot easier if Microsoft had two operating systems, one consumer focused and the other designed for the enterprise. This realization puts a whole new spin on the Windows 8 strategy.

There are two versions of Windows 8 — one a direct descendant of Windows 7 designed for the Intel architecture and a second brand new ‘RT’ — version designed for ARM based devices. Initially this looked to be a retrograde step. The company spent years and billions of dollars trying to collapse it’s complex operating system portfolio down to single re-usable Windows code base. Now the company has gone in the opposite direction. Microsoft now has two versions of Windows; one enterprise focused and one purposefully designed for consumer devices.

Breaking the company up into independent consumer and enterprise focused entities would  drive huge clarity and focus and unlock a lot of shareholder value. Having two distinct versions of Windows 8 aligned to each segment just made that a lot more straightforward.

Evidence and Supposition

Whether Steve and the leadership team ever decide that Microsoft and its shareholders would be better served by a managed breakup of the company remains to be seen. What is clear from recent evidence is that the company is embarking on the biggest change in business strategy in Microsoft’s history. Their ambition is no less than to become a global leader in consumer devices and services.

Microsoft’s leadership team understands that the only way to be successful with this strategy is to take control of the entire end-to-end value chain and to do so in a deeper way than even Apple has achieved. Such a radical strategy would give Microsoft significant margin advantages over its competitors and would enable it to take complete control over the end consumer experience — something it has desperately lacked in its battle with Apple.

The re-alignment of Microsoft’s online services as value enhancers for a device-led strategy will also bring much needed focus to those investments and in doing so would enable Microsoft to take the high-ground on consumer privacy in their battle with Google.

If Microsoft pull this off then the consumer technology business — and the company’s stock — is in for a wild ride.

This weeks Partner’s conference has once again exposed Microsoft’s complete lack of any credible consumer strategy. On the one hand Ballmer claims he’s going to leave no “stone unturned” competing with Apple. Yet on the same day he also states that Microsoft’s own Windows 8 hardware “Surface is just a design point.” Note to Steve: Those are two mutually incompatible objectives.

If Ballmer really is serious about competing with Apple then Microsoft will need to control its own hardware destiny — in pcs, tablets and smartphones — and be completely committed to that strategy. Unfortunately that will also require a willingness to throw the OEM partner community under the bus. Protecting OEMs while aggressively competing with Apple are incompatible strategies. A textbook case of the Innovators dilemma.

Shareholders are likely to pay a very high price if Ballmer continues to believe the fantasy that he can accomplish both these competing objectives and still be successful in the consumer business.