Research Library

A Lifecycle Approach to Cloud Computing – discussion document

{{ Page[0].addThisButton }}

This discussion document launches a new direction in our Organizing IT for the Future research domain

Irrespective of how a person might feel about the hype surrounding cloud computing, the current phenomenon is not only real but also masks a more powerful hidden truth: the way we organize ourselves appears to be flawed. Using the example of cloud computing this discussion document will try to shed light on the organizational economics that are driving cloud computing whilst providing some insight into appropriate future organizational structures.

At the heart of this paper is the notion of competition and of how individuals, companies and governments can find more effective ways of competing. By embracing the underlying complexity of technological change we hope to show how this can be achieved.

The paper itself is derived from a set of experimental practices that were commercially implemented between 2003 and 2006 within a subsidiary of Canon Europe. Those practices, developed through trial and error, led to high rates of innovation and efficiency. However, whilst the practices worked, the theoretical framework to explain why this occurred wasn't developed until much later, in 2007.

By providing the reader with that underlying framework, we hope to foster a deep insight into how both innovation and efficiency can be managed effectively, and to make some key observations about cloud computing. The framework will also provide a starting point for LEF's future research into the space.

"There is nothing more difficult to plan, more doubtful of success nor more dangerous to manage than the creation of a new order of things."
Machiavelli, The Prince (1513).
{{ bp-break }}
{{ Page[0].addThisButton }}


Key observations for cloud computing


For reference purposes, the following summarizes this paper's key messages, to be considered when dealing with cloud computing:

  1. Cloud computing is an inevitable evolution which is rapidly spreading in the marketplace.

  2. Cloud computing is being driven by both the enterprise and consumer ecosystems.

  3. Cloud computing is happening today because of the confluence of four factors: concept, suitability, technology and change in attitude.

  4. Not all activities managed by IT are suitable for the cloud.

  5. Cloud computing can be broken into five distinct categories: Infrastructure, Platform, Software, Process and Orchestration.

  6. Common roles are emerging in the cloud: Provider, Enabler, Broker, Store, Exchange and Assurance.

  7. The benefits and risks of cloud computing are standard to the evolution of any activity.

  8. Care should be taken with many of the assumptions behind the cloud, including: infinite supply, capability, commodity provision, heterogeneous demand, good-enough components and well-ordered markets.

  9. Care should be taken in areas such as increased organizational strain, increased competition through reduced barriers to entry, the legacy question and choice of future standards.

  10. Numerous strategies exist to balance the benefits and risks of cloud computing, including hybrid solutions, use of marketplaces and use of brokers.

  11. Beware the misconceptions of cloud computing, including a reduction in overall IT expenditure, green cloud and virtual data centres as clouds.

  12. Vendor strategies are not always obvious.

  13. The overall consequences of the cloud include increased rates of innovation, disruption to existing vendors and potential loss of barriers to entry.

  14. Loss of barriers to entry may also disrupt mechanisms of control.

  15. The impact of cloud computing is not confined to companies.

  16. The future of cloud computing is likely to be focused on Platform and Process.

{{ bp-break }}
{{ Page[0].addThisButton }}


Organizational economics


An organization consists of many components and ecosystems

A business is a living thing, comprising a network of people, a mass of different activities, and reserves of capital including financial, physical, human and social. It consumes, it produces, it grows and it dies. Like all organisms, any business exists within a number of ecosystems in which it competes and cooperates with others; it is shaped by and shapes its environment, and hence needs to adapt constantly merely to survive1.

People come and go, activities change, and hence all firms are in a constant state of flux. In any industrial ecosystem, new activities (innovations) are a consequence of competition and those that are useful will diffuse throughout the ecosystem becoming more of a commodity. This constant change creates a paradox, identified by Salaman and Storey2:

"Survival requires efficient exploration of current competencies and 'coherence, coordination and stability'; whereas innovation requires discovery and development of new competencies and this requires the loosening and replacement of these erstwhile virtues."

These two extremes of survival (today and tomorrow) have diametrically opposite concerns, and the techniques, tactics and methods needed to manage each are entirely different. Those who manage organizations are therefore caught on the twin horns of a dilemma: how is it possible to be standardized and efficient as well as innovative and new, without prejudicing your survival – either today or tomorrow?

The effects of this on business can be seen in the constant restructuring to cope with new paradigms, and in the yo-yoing of popular management theories between opposites in a scramble to maintain order. A more effective balance can be found through embracing both goals simultaneously. This requires a rethinking of how we organize, and a realization that what really matters is not innovation or efficiency per se, but how we continuously manage the path between the two. To explain why this is the case, we will look at change and how this affects organizations using the example of cloud computing.

  • 1. Leigh Van Valen, A New Evolutionary Law, Evolutionary Theory 1: 1—30, 1973
  • 2. Salaman and Storey, Managers' Theories About the Process of Innovation, Management Studies 39, 147-166, 2002
{{ bp-break }}
{{ Page[0].addThisButton }}


Organizational economics: Diffusion


Innovations diffuse through a society

Everett Rogers3 explained how innovations are communicated over time through various social structures and are consequently either spread through adoption or rejected in a society. The figure above provides a graphical illustration of this process, showing the cumulative adoption of an innovation over time until saturation is reached, with each stage governed by different social groups (for example, innovators, early adopters, early majority). Several considerations to this concept need to be recognized:

  • The rate of diffusion is not constant – Comparisons over time provide a wide range of adoption curves and a general observation that the diffusion of innovations is accelerating.

  • Not all innovation spreads – Even where an innovation has utility (usefulness), a number of factors can influence its adoption. As Geoffrey Moore noted4, there is a chasm between the early adopters of an innovation and the early majority.

  • Diffusion is not continuous – Highlighted by Christensen's work on disruptive innovation5, diffusion of one innovation can be disrupted by the introduction of a new technology that offers different performance attributes from those established in existing value networks.

  • Diffusion consists of multiple waves – Innovations tend to spread through waves of improved versions. In the early stages of a technological change, this rate of improvement tends to be slow and then accelerates until reaching a more mature and slow improving stage6. One consequence of the diffusion and maturing of a technological innovation is that increased information about the technology reduces uncertainty7 about the change. Each improved version increasingly adds definition, eventually providing a system that can be considered featurecomplete, mature and generally well understood.

  • 3. Everett Rogers, Diffusion of Innovations, Free Press, 4th Edition, 1995
  • 4. Geoffrey A. Moore, Crossing the Chasm, Harper, 1991
  • 5. Clayton M. Christensen, The Innovator's Dilemma, Harvard Business Press, 1997
  • 6. D. Sahal, Patterns of Technology Innovation, Addison Wesley, 1981
  • 7. E. M. Rogers and D. L. Kincaid, Communication Networks: Towards a New Paradigm of Research, Free Press, 1981
{{ bp-break }}
{{ Page[0].addThisButton }}


Organizational economics: Uncertainty


Uncertainty is the key to understanding change

The standard model that we've just outlined contains different rates of diffusion, multiple waves of improved versions and both sustaining and disruptive innovation. Whilst on the surface it appears simple, the pattern is highly complex, with extensive debate over what actually constitutes an innovation. For example, whether something is an innovation or not depends upon the perspective one is adopting8. Hence, the Bessemer converter was a process improvement to iron and steel manufacturers but a product innovation to suppliers of equipment for those industries. Equally, the modern-day provision of computing resources through large utility providers (such as Amazon's EC2 service) is certainly an innovation for those service providers, but for consumers the use of computing resources in business is not9.

In order to create some form of basis for investigation, Fagerberg's10 definition of innovation as the "first attempt to put an idea into practice" will be used. Hence the innovation of the phone will be considered to be the first phone, not a product improvement or feature differentiation. From this starting point, the figure above maps the history of three activities across two axes – the ubiquity of that activity (that is, a measure of how common an activity is) against its certainty (that is, a measure of how well defi ned and understood an activity is)11. The activities cover different time periods and include televisions and their use between 1935 and 1982, VCrs (1984-2002) and telephones and their use (1951-2000).

The graph hypothesizes a simple S-curve relationship between the ubiquity and certainty of an activity, and demonstrates a pathway for how a rare and poorly understood innovation transitions to a more common and well-defined commodity.

  • 8. Simon Kuznets, Innovations and Adjustments in Economic Growth, Swedish Journal of Economics, 1972
  • 9. Caminer et al, LEO: The Incredible Story of the World's First Business Computer, 1998
  • 10. J. Fagerberg, Oxford Handbook of Innovation, OUP, 2005
  • 11. Private research, 2006-2009
{{ bp-break }}
{{ Page[0].addThisButton }}


Organizational economics: Lifecycle


Activities transition through a commonly repeated pattern

Through examination of a wide range of business activities, a commonly repeated pattern emerges: innovation, custom-built examples, productization and eventually commodity provision. These 'domains' have been sketched onto the figure above along with two provisioning models – services and utility services.

This pattern of change is commonly referred to as commoditization and examples of this can be seen in many industries from pharmaceuticals to music to finance. It should be noted that:

  • Not everything becomes ubiquitous – physical constraints, the impact of consumption, the usefulness of something, further technological change, redundancy, socio-economic, political and geographical factors and so on can hinder if not halt the spread of an activity.

  • Economic forces drive the pattern – These are user competition (that is, the constant demand for any new activity which is a source of advantage) and supplier competition (that is, the constant improvement of an activity – a better phone, a better computer etc.).

  • There are accelerators – Whilst the pattern itself is not a time-based sequence, there are a number of factors that alter the speed at which an activity evolves through its lifecycle. Common accelerators include network effects, social connections12 and methods of increasing participation. Common decelerators include branding, complexity and confusion of choice. Patents and other means of intellectual property protection can both accelerate and decelerate the evolution of an activity depending upon the length of their term and the likelihood of independent discovery, a discussion of which is beyond the scope of this paper.

  • 12. Amin Saberi, Spontaneous Emergence of Influence in Online Systems, PNAS, 2010
{{ bp-break }}
{{ Page[0].addThisButton }}


Lifecycle: Cloud


Cloud computing simply represents an evolution of activities

Cloud computing is a marketing term used to both describe and mask a real change that was first predicted in Parkhill's 1966 book, The Challenge of the Computer Utility.13

Parkhill asserted that in the future, computing would be provided in the same manner as electricity – through utility services. These services would be online, charged on a basis of consumption and offer an elastic supply of standardized resources from raw compute to discrete applications. His view was derived from examination of the evolution of other industries and consequently he speculated that we would see the formation of public, private, government and community computer utilities with organizations often combining several of these sources. His predictions map precisely to the modern NIST14 definition of cloud computing and the term can be considered synonymous with 'computer utility'.

An essential requirement for a computer utility is that the activity provided must have evolved to a point where it is both ubiquitous and well defined enough to support the volume operations needed for a utility. Many activities described by IT have now reached that stage – for example, the above shows the evolution of one activity, the use of computing infrastructure. It should be noted that not all IT activities are suitable for provision by computer utilities and that there is a significant difference between those activities that will be provided as computer utilities and those that will be built upon them.

The modern-day confusion over the cloud is often associated with the wide diversity of activities being affected, the transition from one state (of products) to another (of utility services) and the lack of an obvious path for evolution. For example, whilst server infrastructure has mainly commoditized and evolved within the business ecosystem, activities such as provision of email have been commoditized in the consumer space.

  • 13. D. Parkhill, The Challenge of the Computer Utility, Addison Wesley, 1966
  • 14. Definition of Cloud Computing, version 15, National Institute of Science & Technology
{{ bp-break }}
{{ Page[0].addThisButton }}


Cloud: Why now?


Cloud computing is due to multiple factors

Whilst Parkhill prophesied this change 44 years ago, the evolution of any activity from its provision through products to utility services requires more than just concept before it can occur. The other necessary factors, which exist today, include:

  • Suitability – The activities offered need to be widespread and well defined. The term 'as a service' is a misnomer for 'as a utility service', since the ability to provide IT as a service has long existed. Many earlier services models (such as application service providers or ASP) offered highly customized environments, but it is the standardized, non-differential nature of utility provision that separates cloud computing from these earlier models.

  • Technology – Virtualization15 is one of many enabling technologies. It is often described as essential to cloud computing despite its lack of use by many computing utilities. Whilst virtualization enables more efficient provision of infrastructure resources, it should be noted that a utility has other fundamental requirements including self-service, multi-tenancy, charging and delivery mechanisms.

  • Attitude – The largest factor in the drive towards utility provision has been the changing relationship between IT and the business, which originated with Strassmann's work16 on the lack of correlation between IT spending and value. This was reinforced with nicholas Carr's paper,17 which showed that as activities described by IT became more ubiquitous they had diminishing strategic value and were simply a cost of doing business suited for utility provision. Whilst these works have changed the mindset of businesses everywhere, it is unfortunate that the provocative title of Carr's article and his later book Does IT Matter? have often been misinterpreted to mean the whole of IT rather than discrete activities within it.

  • 15. Popek, Formal Requirements for Virtualizable Third Generation Architectures, ACM, 1974
  • 16. Paul Strassmann, The Value of Computers, Information and Knowledge, 1996
  • 17. N. Carr, IT Doesn't Matter, HBR, May 2003
{{ bp-break }}
{{ Page[0].addThisButton }}


Cloud: Market


Cloud computing can be characterized into discrete markets

The industry that is developing around cloud computing is commonly sub-divided into five distinct categories: Infrastructure, Platform, Software, Process and Orchestration. Each category has three primary markets: end consumers, service providers and technology vendors, as summarized in the figure above. For example, Amazon EC2 is a service provider offering infrastructure targeted towards end consumers, whereas opscode is a vendor offering configuration management technology for private and public clouds.

Whilst the size of the market is complicated by different terminology, assumptions and biases, the following can be noted:

  • Predictions for the size of the cloud computing market vary between $25 billion18 and $150 billion19 in 2013, covering an estimated 20% to 65% of all workloads by 201520.

  • Surveys of large enterprises (2,500+ employees) show adoption rates between 15%21 and 77%,22 with up to 34% of companies using hybrid (public and private) cloud computing.

The real issue for any company is not the general market adoption of a technology but its adoption by its competitors. Such information is specific to an industry, the company and the activities undertaken. Forecasts for growth can only be seen as a relative guide to importance of cloud computing in general and the variance in figures is a symptom of treating a complex market with a myriad of defi nitions as though it was a single entity.

  • 18. Cloud Computing – SaaS, PaaS and IaaS, Renub Research, 2009
  • 19. Sizing the Cloud, Gartner Forecast, March 2009
  • 20. Marten Mickos (Eucalyptus), Review of analyst forecasts presented at the OSCON Cloud Summit 2010
  • 21. Cloud Computing Survey, ITC, 2009
  • 22. Novell, presentation at VMWorld; own research and survey, Oct 2010
{{ bp-break }}
{{ Page[0].addThisButton }}


Cloud: Emerging roles


Patterns in the cloud computing market are emerging

In the cloud computing space there are more than 180 combinations of category and market, with over 300 identifiable players. Common emerging roles include:

  • Provider – offers direct services, for example, Amazon, rackspace, google, Microsoft and Salesforce.

  • Enabler – offers technology to develop, build, manage or automate cloud services whether public facing or internal private clouds. Examples include VMWare, Eucalyptus, rightScale, openStack, Enstratus and Opscode.

  • Broker – The term 'broker' has been co-opted by many companies to describe systems that are little more than management mash-ups and bridges between different services. There are, however, a few examples of companies acting as genuine brokers (intermediaries), including CSC's trusted cloud.

  • Store – offers everything from virtual machine images for infrastructure services to discrete cloud applications, for example, CohesiveFT and rpath.

  • Exchange – Enables the direct selling, re-selling and purchase of computer resources, for example, SpotMarket and Zimory.

  • Assurance – An essential requirement of a well-functioning market is assurance information on the providers. Whilst there are early attempts to develop a mechanism for assurance (for example, CloudAudit group) and monitoring (for example, CloudMarketWatch), there currently exists no established rating agency for the cloud.

Given the complexity of the market and how roles are only starting to emerge, the tendency for any organization might be to wait until there is greater clarity. However, all organizations must consider that as with the evolution of any activity, cloud computing can create real and significant benefits right from the start.

{{ bp-break }}
{{ Page[0].addThisButton }}


Lifecycle: Benefits


Evolution of activities creates organizational benefits

As an activity evolves it creates numerous organizational benefits, including:

  • Greater links between expenditure and consumption – The classic example of this is utility charging for commodity services, which enables the conversion of capital to operational expenditure, a reduction of working capital tied up in unused assets, and the ability to scale up and down according to variable demand (for example, due to seasonal demand, transitory use, the success of a project or its failure). Moreover, increased transparency of consumption enables identification and improvements of inefficiencies within a system.

  • Increased focus – Through reduction in capacity planning efforts and the capability to outsource non-core activities.

  • Increased efficiencies – From economies of scale achieved through volumeoperations and maximization of utilization rates across heterogeneous demand.

  • Increased agility and rates of innovation – Through self-service, automated provisioning and use of standardized components.

Whilst many of these benefits might indicate a potential reduction in expenditure, the introduction of a more commoditized means of providing an activity often leads to an increase in consumption, particularly if the activity becomes a component for the provision of other activities. For example, the commoditization of electricity provision resulted in not only increased efficiencies but also an explosion in consumption, as new activities such as consumer electronics developed23. This effect is known as Jevons' paradox24, the constituent drivers of which are price elasticity, a long tail of unmet demand, co-evolution and increased rates of innovation.

  • 23. Historical World Electricity Consumption, IEA, 2002
  • 24. William S. Jevons, The Coal Question, Macmillan & Co, 1865
{{ bp-break }}
{{ Page[0].addThisButton }}


Cloud: Benefits


The benefits of cloud computing are similar to other utility markets


The benefits of cloud computing are the same standard benefits that apply to any activity that evolves along its lifecycle, whether telecoms, banking or electricity provision: increased focus, economies of scale, utility charging and increased agility. A detailed examination of this is provided in the LEF report Doing Business in the Cloud25.

It should be noted that many of the commonly cited benefits are based upon assumptions that have limitations. For example:

  • An assumption of infinite supply – All providers have some level of constraint and there is no guarantee that further requests for compute resources can be met.

  • An assumption of capability – Whilst Capex to Opex conversion sounds attractive, an assumption is made that a company has the internal procedures capable of dealing with variable pricing. It is considered likely that many new services will run into conflict with internal budget procedures.

  • An assumption of commodity provision and heterogeneous demand – Corporate data centres have tended towards low utilization rates to cope with unexpected or seasonal demands, low MTTR (mean time to recovery) for replacement of physical equipment, or transitory requirements (that is, test environments). Here the provision of infrastructure through compute utilities may be beneficial. However, an assumption is made that changes in demand will balance over an entire population (that is, that it is heterogeneous), which discounts the risk of an unexpected systemic effect (a black swan event26).

    A black swan event could result in a temporary denial of new requests for compute resources as demand increases to exceed supply. However, a black swan event could also result in some providers being unable to deliver what has already been promised, due to oversubscription (also called overbooking or thin provisioning). Oversubscription is used to sell more compute resources than are available, with the provider gambling on a heterogeneous demand in order to deliver the service to specific customers when they need it. These providers may be unable to meet the demands of existing customers should too many of them need to fully utilize services they've paid for at the same time.

  • An assumption of well-ordered markets – Without standardization and easy switching between providers, the marketplace cannot be considered well ordered with effective pricing competition. Hence, many of the achievable economies of scale are unlikely to be passed on to the consumers. The use of oversubscription, the emphasis placed on the need for high SLAs, the lack of effective comparison and the introduction of configurable environments serve to fragment the market to the disadvantage of consumers.

  • An assumption of 'good enough' components – Whilst the economies of scale are tangible, they are based upon the provision through volume operations of standardized 'good enough' components (that is, noncustomized commodities). A consumer should therefore consider that whilst the components are low cost and with low MTTR, they are also inflexible (being standard units for a provider).

  • 25. D ouglas Neal et al, Doing Business in the Cloud: Implications for Cost, Agility and Innovation, LEF, August 2009
  • 26. Nassim Nicholas Taleb, The Black Swan: The Impact of the Highly Improbable, 2007
{{ bp-break }}
{{ Page[0].addThisButton }}


Lifecycle and cloud: Innovation


Innovation and commoditization are part of a virtuous circle

As an activity becomes more of a commodity it has a declining differential value (due to its ubiquity). This reduction in value enables new activities to arise, as the resource and attention of companies focus towards sustaining future value. Joseph Schumpeter described27 this concept as "creative destruction".

However, commoditization also accelerates innovation through a concept known as componentization28. Formally this describes how the speed of evolution of any system is directly related to the organization of its subsystems – the 'building on the shoulders of giants' phenomenon. Today, application development is fast because of stable development environments built upon more layers of stable components. If these weren't stable, every application would start with designing a CPU.

From the figure above, as activities evolve driven by user and supply competition to become more stable commodities, they enable (via creative destruction) and accelerate (via componentization) the innovation of new activities. The net result is often a shift in focus towards higher-order systems with the original activity becoming a less visible but essential component.

An example is Maudslay's screw-cutting lathe and the introduction of standardized, mass produced and interchangeable mechanical parts29 (for example, nuts and bolts), which resulted in not only a rapid growth of innovation during the industrial revolution but also the consignment of these parts to less visible but essential subsystems of grander machines. In the same manner, cloud computing should accelerate innovation of higher-order systems (mash-ups are a common example) whilst becoming a less visible but no less essential subcomponent.

  • 27. Joseph Schumpeter, Capitalism, Socialism and Democracy, 1942
  • 28. Herbert Simon, The Architecture of Complexity, American Philosophical Society, Vol 106, 1962
  • 29. L.T.C. Rolt, Great Engineers, G. Bell and Sons Ltd, 1962
{{ bp-break }}
{{ Page[0].addThisButton }}


Lifecyle and cloud: Risks


Evolution of activities creates organizational risks

Three main categories of risk are associated with the evolution of any activity. These include disruption risks related to the change away from a previous practice, transitional risks related to the adoption of a new practice, and outsourcing risks related to the use of a third-party product or service:

  • Disruption risks include changing business relationships, loss of previous investments or political capital, reducing barriers to entry, and increased competition in an industry.

  • Transitional risks include confusion over the new methods of providing the activity, concerns over transparency and trust with new providers, concerns over governance, and issues over security of supply.

  • Outsourcing risks include suitability of the activity for provision by a third party, existence of pricing competition, lack of second sourcing options or switching between alternatives, and loss of strategic control through dependency on vendor or service provider.

The risks of cloud computing are identical. However, the following are worth emphasizing:

  • The legacy question – Many environments will need to be re-designed for the cloud and hence care should be taken to examine whether 'legacy' systems provide some form of advantage or whether more commoditized means of achieving the same results exist.

  • The problem with choice – Selection of the future standard is critical but identifi cation of that standard is difficult. For example, enterprises heavily invested in IPX/SPX until forced to adopt the public de facto standard of TCP/IP, incurring high transition costs.

  • Organizational strain – Commodity IT activities are suited to management methods that promote efficiency, but equally commoditization increases the rate of innovation. The effective exploitation of this effect requires a polar opposite set of management methods. By increasing the emphasis on both efficiency and innovation, the cloud is likely to create strain within organizations that tend to use one management practice across all of IT.

{{ bp-break }}
{{ Page[0].addThisButton }}


Lifecycle and cloud: Bundling


Examining organizational activities is non-trivial

When examining activities within an organization, there are a number of pitfalls to be avoided. As noted previously, innovation is considered to be the first time a new activity is introduced, but many differentiations of products are falsely described as innovations. Care should therefore be taken to avoid the supplier's perspective.

Unfortunately there exist numerous techniques that are used to alter a user's perspective, including branding, confusion of choice and complexity. For example, within the consumer ecosystem, whilst the vacuum cleaner can be considered fairly well-defined and ubiquitous, by associating its offering with notions of lifestyle choice, giving a plethora of options (different versions designed for specific purposes) and using visibly overt engineering (complexity), Dyson has disassociated the act from the device and successfully created a status symbol30. Care should be taken to treat cloud computing as a commodity and to discount supplier claims appropriately.

Care must also be taken over the question of bundling; many products today represent a plurality of activities – for example, the iPhone (see the figure above) is more than just mobile telephony. In the case of cloud computing, a primary cause of confusion is the bundling of the act of using computers with architectural principles for implementation.

It is important to understand that many of the benefits of computer utility provision (that is, cloud computing) are only available through use of modern architectural principles (for example, horizontal scaling, design for failure). That is to say, moving a legacy environment using legacy architectural principles (for example, vertical scaling, N+1) to the cloud will have limited benefits. Care must be taken to unbundle benefits associated with computer utility provision (for example, economies of scale, increased agility) from benefits associated with architectural choice (for example, scaling, resilience) to avoid unrealistic expectations of the cloud. A legacy application suffering from vertical scaling issues is unlikely to find resolution by simply moving to an infrastructure cloud.

  • 30. Michael Schrage, The Myths of Commoditization, MIT Sloan Management Review, 2007
{{ bp-break }}
{{ Page[0].addThisButton }}


Lifecycle and cloud: Ecosystems


The business ecosystem does not function in isolation

The evolution of an activity is complicated by the existence of multiple ecosystems. Activities can exist in, transfer between and commoditize within any ecosystem. For example, whilst CRM has mainly commoditized in the enterprise space, email has crossed over to the consumer space and has been more aggressively commoditized. Today, these heavily commoditized services are now transferring back into the enterprise ecosystem.

The key issue to consider is in which ecosystem the evolution of an activity is governed. The process by which this governance transfers to the consumer ecosystem was first described by the LEF as consumerization31 and email is just one example of this. The consumerization of technology and its subsequent commoditization in the consumer ecosystem can have a profound effect on the operation of enterprises using that technology. It should be noted:

  • Not everything transfers.

  • Barriers to entry can erode quickly – The impact on the media industry of the commoditization of the means of mass communication occurred primarily in the consumer space over a single decade.

  • Ecosystems are not isolated – people are part of multiple ecosystems, hence activities in the enterprise (which may be provided through products) tend to start with custom-built solutions and hobbyist equivalents in the consumer space.

  • The rate of commoditization is not uniform between ecosystems.

Numerous IT activities have undergone consumerization-creating responses such as BYOT (bring your own technology) and shadow IT. All organizations should therefore examine the consumer space for highly commoditized services and consider whether barriers to entry into their own industry will reduce. In many cases, adoption of the service and creation of higher-order systems can be used to re-establish those barriers.

  • 31. Douglas Neal and John Taylor, LEF, 2001
{{ bp-break }}
{{ Page[0].addThisButton }}


Lifecycle: Consequences


Evolution has unintended consequences

Contrary to popular belief, most organizations have little choice over whether to adopt an evolutionary change. As an activity evolves, the operational benefit for competitors creates an increasing organizational pressure for adoption. Shown in the figure, this is the organizational equivalent of the red Queen's Hypothesis32: the constant need to evolve in order to stand still relative to a surrounding competing environment.

Furthermore, commoditization also creates and is affected by co-evolution through componentization. As an activity becomes a component within another activity, the growth of the latter can drive the commoditization of the former. An example of this can be seen in the evolution of the electronic switch, from Fleming's valve to complex products containing multiple switches (for example, the Intel 4004) and the development of new industries (for example, digital computers) based upon these components. These new industries drove further commoditization of switching and as a consequence created ever more powerful components (containing millions of switches), which in turn enabled further industries (for example, digital phones). Componentization and co-evolution can be powerful forces for future innovation.

Lastly, commoditization can impact control within an organization. Newspapers once held a privileged position due to their access to expensive physical capital assets, such as printing presses and distribution systems, which were essential to mass communication. Access to these assets limited the number of competitors and was also a significant point of control; if a person wanted to be a journalist then they had to work for a company with access to those assets. Aggressive commoditization of the means of mass communication (through the internet and digitization) undermined not only the barriers of entry into the media industry, encouraging the formation of alternative news channels, but also the points of control. Today's journalists can build their own news channels33.

{{ bp-break }}
{{ Page[0].addThisButton }}


Cloud: Consequences


Cloud consequences include disruption, innovation and reducing barriers to entry

As activities shift from a product to a utility services world, feature differentiation is replaced with service differentiation – the enabling technology becomes less important than the service itself. Open source is therefore likely to become the most dominant model in the cloud particularly as it provides a route to creating a competitive market of service providers. This shift, shown in the diagram above, has precedents in the replacement of IPX/SPX with TCP/IP and is likely to be disruptive to existing proprietary vendors. Equally, the change is likely to be disruptive to hosting providers (service model) that cannot adapt to a commodity environment (utility service model).

The consequences of cloud computing can be summarized as disruption to existing models, reduced barriers to entry in orthogonal industries (including loss of control) and acceleration of innovation within IT.

These effects are not limited to the competition between companies. According to a recent CEBR report, the cumulative economic benefits of the cloud within France, germany, Italy, Spain and the UK are estimated to be 750 billion Euros by 2015 across manufacturing, distribution, retail, banking, government, health and education34. These consequences are a probable driving force behind the Chinese government's announcement that it aims to become "the largest cloud computing centre in the world"35. It should be noted that whilst the technology may shift from closed (proprietary) to open with the formation of competitive markets, this does not necessarily mean an open ecosystem of providers. Radio broadcasting transformed from proprietary technology with millions of consumer broadcasters to open technology and licensed operators through government regulation. Many industries have undergone this 'closed but open' to 'open but closed' transition36.

  • 34. The Cloud Dividend, Centre for Economic & Business research, Nov 2010
  • 35. Cloud Computing on the Far Horizon, Caixin Online, Nov 2010
  • 36. Timothy Wu, The Master Switch: The Rise and Fall of Information Empires, Nov 2010
{{ bp-break }}
{{ Page[0].addThisButton }}


Cloud: Managing


Numerous strategies exist to balance the benefits and risks of the cloud

The Red Queen Hypothesis shows us that the use of cloud computing is not a question of if but when. Therefore, any organization must examine how to adopt the cloud and maximize the benefi ts against risks. Fortunately, we can learn from successful strategies that have been applied to other utilities and are entirely valid for cloud computing:

  • The Hybrid – Industrial consumption of electricity usually involved a mix of both public and private sources of generation. Even today, the modern data centre will combine multiple public supplies with back-up generators, UPS or even self-generation. This mixing of both public and private is known as a hybrid model and it can be effectively used to mitigate transitional and outsourcing risks (most notably security of supply, governance and loss of strategic control) but at the cost of reducing benefits (most notably economies of scale).

  • The Use of Marketplaces – Almost all of the outsourcing risks (lack of pricing competition, lack of second source options, etc.) can be mitigated through the use of a competitive marketplace consisting of multiple providers and easy switching between them. There are numerous attempts currently under way to create such marketplaces, most notably the OpenStack37 effort supported by a wide consortium of companies.

    The principal requirements of a marketplace include standards (for access to code and data), common reference models (to ensure semantic interoperability), assurance (to ensure compliance with the standards) and minimal exit costs from providers. Care must be taken when choosing a marketplace to consider the issue of strategic control and whether one vendor will exert control over the entire market (known as a controlled market). Use of open source technology is preferable since no single company can exert control, however this does not preclude government legislation for licensed operators.

  • The Use of Brokers – Acting as a trusted intermediary between the client and the world of suppliers. Both brokers and clearing houses38 are starting to emerge in the cloud space.

{{ bp-break }}
{{ Page[0].addThisButton }}


Cloud: Standard misconceptions


With most changes, various myths and misconceptions develop during the transition from one mode of operation to another, and cloud computing is no exception. Consider the following frequently heard claims:

  • The cloud will save you money/be green – Whilst there are numerous reports on the financial and green benefits39 of the cloud, they typically ignore demand-side effects. The provision of activities through ever more efficient means (in terms of both cost and self-service) will result in increasing consumption (the Jevons paradox). Hence whilst cloud computing may offer advantages on a per unit basis, when examining entire IT expenditure, we're unlikely to see any significant long-term reductions.

  • The cloud is an innovation – But only to service providers. Care should be taken to avoid considering the cloud as some form of innovation as this can easily lead to mis-management.

  • My virtual data centre is a cloud – Many virtual data centres are built for purpose (that is, for specific applications) and have arisen through maximizing the utilization of non-commodity hardware. However, cloud services are all about commodity provision and are designed for volume operations. Care should be taken as the economics are widely different and many 'private clouds' are likely to be highly uncompetitive40 with their public counterparts. Direct comparison is also complicated since many internal technology infrastructure projects don't take into account the true costs of ownership41.

  • Cloud is about infrastructure and software – At the moment, there is a high degree of emphasis on both cloud infrastructure and software providers. Going forward, both the infrastructure and software levels of cloud computing are likely to become stable underlying components of higher-order process and platform systems.

  • Strategies are clear – There are many marketing strategies being deployed. Common examples include: be the standard; an open market; and cloud but like your data centre and trusted. However, the intentions of vendors are far from clear.

For example, at first reading Vmware might be seen as focused on being a technology enabler for infrastructure services by offering proprietary technology that extends your data centre into the cloud. However, an examination of the acquisition strategy of Vmware would show that it is almost exclusively focused on open source platform technology. Given its existing revenue streams and the potential disruption in the infrastructure space through open source technology such as KVM, the company might appear to be hedging its bets. However, in recent press interviews it has become clear that Vmware is positioning itself towards Platform-as-a-Service computing42 with server virtualization seen as having limited long-term value in the cloud.

  • 39. Cloud Computing for Business Users, Microsoft Research Study, 2010
  • 40. The Economics of Cloud, Microsoft, Nov 2010
  • 41. David Katz, CIOs Called Clueless Over Costs, CFO Magazine interview, Sep 2010
  • 42. VMware Knows the Cloud Doesn't Need Server Virtualization, GigaOm interview, July 2010
{{ bp-break }}
{{ Page[0].addThisButton }}


Lifecycle: Macro


Commoditization of one activity can accelerate the evolution of all

In 1925, Nikolai Kondratiev43 first proposed the existence of supercycles of economic development. Perez44 characterized these cycles into waves clustered around technological and economic paradigm shifts. For example, the Industrial Revolution (the first wave) included factory production, mechanization, transportation and development of local networks whereas the Age of Oil and Mass Production (the fourth wave, after the ages of Steam and Electricity) included standardization of products, economies of scale, synthetic materials, centralization and national power systems.

Each of these waves is centred on a major cluster of innovations, with each innovation undergoing a cycle of incremental improvement until it reaches a plateau of diminishing returns with widespread diffusion of the new paradigms (that is, it becomes commoditized). Whilst the model is a time-based sequence, it directly maps to lifecycle.

Taking a lifecycle perspective can shed some further light on these changes. First, the impact of componentization is profound not only in a current economic wave but also in future economic waves. Second, certain types of technological and social change can impact the overall rate of diffusion and hence the evolution of all change. For example, the printing press, telephony, open source, the internet and social networks all accelerated the general rate of evolution of all activities by increasing communication or participation.

Whilst cloud computing will become a major component of future technological and economic waves by allowing a faster rate of deployment of new innovations (through componentization), it is unlikely to accelerate the overall rate of evolution because utility computing (unlike the internet) does not inherently provide a better means of communication, dissemination or participation.

  • 43. N. Kondratiev, The Major Economic Cycles, 1925
  • 44. C. Perez, Finance and Technical Change, Judge's Institute, 2004
{{ bp-break }}
{{ Page[0].addThisButton }}


Lifecycle: Profile


Examining activities within a company enables a shape to be determined

An organization consists of many activities at different stages of evolution. By plotting frequency of activities against their stage of lifecycle, a profile for an organization can be determined as shown in the figure above. New activities are constantly appearing, being commoditized through user and supply competition, and thus enabling further innovation.

The shape of a company is transitory; however, it is the LEF's position that it's not just companies that have shape but industries too. The profile of the high-tech web industry, where emphasis is placed upon highly innovative activities, is unlikely to match the profile of the mining industry, where emphasis is placed upon operation of well-defined and understood processes. This is not to say that one industry is innovative and the other not, merely that the balance will be different between the two. This concept is important for several reasons:


  • What matters is your organization's shape and its relationship to the industry shape – If the profile of a company is skewed towards more innovation than its industry, this could mean either that the company is investing in innovation (attempting to achieve a source of advantage) or that it is treating commodity activities as though they were innovations (a source of competitive inefficiency). An example of this was highlighted at a recent forum of 60 CIos, where all attendees identified that CRM was an activity they undertook but which provided little or no differential advantage between them. despite it being more of a commodity, a majority were spending vast amounts of resources customizing this activity and treating it as a form of innovation.

  • Techniques applied to one industry may not be relevant to another – Whether we are discussing organizational structure or management methodologies, consideration must be given to the applicability of such techniques across different profiles. For example, a mining company with a profile heavily skewed towards commodities may be more suited to highly structured, linear-type techniques. Alternatively a web 2.0 company, with a profile heavily skewed towards innovative activities, may be more suited to highly collaborative networks and dynamic techniques. Clearly, one size does not fit all.

{{ bp-break }}
{{ Page[0].addThisButton }}


Profile: Management


Activities change properties as they evolve

As an activity moves through the profile, its properties change. For example, an innovative activity is dynamic, deviates from what has been, is highly uncertain, has emerging characteristics, and is a potential source of future worth and competitive advantage with serendipity playing a significant role. It is best described as chaotic by nature. A commodity on the other hand is repeated throughout an industry, standardized, well understood, predictable, a cost of doing business and linear by nature.

A single activity, such as CRM, has evolved through each of these stages, from the early list-based systems of the 1980s, to database marketing and the first products, until eventually utility services such as Salesforce appeared. As it evolved from one state to another its properties changed. This kind of evolution profoundly impacts how an activity should be governed. Taking the simple case of project management, techniques such as Agile development are suited to innovative pursuits because they effectively encourage deviation. For the same reasons these methods are ineffective at governing commodities when compared to structured methods such as Six Sigma, because there deviation is undesirable.

The Cynefin45 framework provides an analogy to this transition with the classification of knowledge systems from chaotic, where there is no understood relationship between cause and effect (hence novel practice can be discovered), to simple (linear) where relationships are obvious to all and best practice can be applied.

A consequence of changing properties is that there cannot be one method or technique that is ideally suited to managing both extremes; a plurality of methods is required. Correctly balancing this matters because innovative activities are a company's source of future worth (creative destruction) whilst commodities are a source of operational efficiency (the Red Queen Hypothesis). It is this that creates the innovation paradox highlighted by Salaman and Storey46.

  • 45. D. Snowden, Cynefin, an Ecological Approach to Sense Making, KMAC, University of Aston, 2000
  • 46. Salaman and Storey, Managers' Theories About the Process of Innovation, Management Studies, 39, 147-166, 2002
{{ bp-break }}
{{ Page[0].addThisButton }}


Profile: Consequences


Failing to consider lifecycle creates alignment and other issues

Most organizations group activities by type (such as IT, Finance, Marketing). Each of those groups has a profile; for example, IT is not one activity but a mass of activities from innovation to commodity. The mismanagement of activities in one group will cause misalignment with others. In the forum of 60 CIos discussed above, the reason why companies were spending vast sums on customizing CRM, an activity which often has little or no differential value, was not a failure of IT or its inability to be efficient but instead a failure of the business to treat that which is principally a commodity as a commodity.

Equally, outsourcing, a successful tactic when applied to commodity activities, is fraught with danger when applied across activity types. Typically, the benefits of outsourcing (for example, economies of scale through volume operations) are only achievable for commodities. But when a type is outsourced this covers an entire profile of activities from the innovative to the commodity. Economies of scale are not achievable with innovative activities and the result is often higher than expected costs (because the activity is not suitable) combined with a loss of strategic control.

Another commonly occurring scenario is the yo-yo between innovation and efficiency. In the most classic example of this, new start-ups tend to have profiles highly skewed towards innovation. As the company grows, efficiency tends to become the watchword due to competition. Standard structured procedures are often introduced but unfortunately these often spread too far, covering not only commodity activities (where they are desirable) but also innovative activities (where they are not). Too often the net result is a management yo-yo between demands for greater efficiency followed by demands for innovation.

A common response to this is an ideas process, an attempt to define that which is uncertain, serendipitous and dynamic. Whilst product improvement of an existing activity (which is in transition) is suitable for such an approach because information (customer feedback) is available, it is not possible to pre-determine the success of genuine innovations (due to their inherent uncertainty). The future is an information barrier and the strategic value of something is inversely proportional to the certainty we have over it – business always contains an inevitable element of gamble.

{{ bp-break }}
{{ Page[0].addThisButton }}


Profile: New patterns


Alternative patterns to structure and management are emerging

Since all business activities undergo evolution from innovation to commodity and the properties of those activities change during the transition, it poses the question of whether organization and management by type actually make sense. Alternative patterns for managing activities appear to be emerging, whether deliberate or accidental. An example of this is the Innovate, Leverage, Commoditize pattern (ILC47) shown above.

Under ILC, different techniques are used to encourage innovation, to leverage any surrounding ecosystem of the business to identify new successful activities (in effect, crowd sourcing) and to commoditize those activities identified to provide core services and enable the next wave of innovation (componentization). Examples of this emergent pattern can be found in Google, Amazon and Salesforce.

A more advanced pattern is to structure a company by lifecycle, with groups such as 'pioneers' (Innovate), 'settlers' (Leverage) and 'town planners' (Commoditize). Here activities flow from one group to another and a cycle of improvement is created. Surprisingly, the critical controlling element in such a structure is neither the rate of innovation nor efficiency, but the role of settlers and how effectively transition is managed.

The major challenge with these models is that they increase the level of management complexity in accordance with Ashby's law of requisite variety48 – that is, the controlling element must have as much complexity as that which is being controlled. Despite the reality of what is being managed, the act of treating a business as consisting of linear, simple, factory-like processes allows for vastly simpler management (albeit at the cost of effectiveness). This situation is acceptable only if all competitors apply the same approach.

The LEF's opinion is that the 'one size fits all' model of management will be increasingly challenged by emerging patterns that consider the true nature of what is being managed.

  • 47. Mark Thompson, Simon Wardley et al., Better for Less, How to Make Government IT Deliver Savings, PBAge, 2010
  • 48. W. Ross Ashby, An introduction to Cybernetics, Chapman & Hall, 1956
{{ bp-break }}
{{ Page[0].addThisButton }}


Organizational economics: Key


To summarize this section: in order to effectively manage activities within an organization, an understanding of what activities a business undertakes, how they are provided, what stage of lifecycle they exist at and what activities are bundled together, is required.

The following should be noted:

  • An organization consists of a mass of people and activities.

  • An organization competes in one or more ecosystem(s).

  • Activities have a lifecycle from innovation to commodity.

    This is due to competition in that ecosystem, through demand and supply effects.

    It should be noted that an activity can spread from one ecosystem to another, which creates the effect known as consumerization.

  • As an activity evolves, benefits and risks change.

    This is why no organization can stand still and is the cause of the Red Queen Hypothesis.

  • As an activity evolves, it enables new innovation.

    The effects of creative destruction and componentization create a constantly accelerating cycle of higher-order systems, which is why life seems to be getting faster.

  • As an activity evolves, its properties change.

    This is why there is no single method or tactic to managing all activities, which in turn creates the innovation paradox.

  • Many modern management failures are a failure to consider evolution.

    This why the outsourcing of activities by type can have very mixed results and is also a root cause of alignment issues within organizations.

  • An organization has a shape.

    Whilst activities in an organization are transitory, a shape for the organization can be determined. Understanding shape is critical for determining what methodologies should be applied where and how to structure yourself.

  • What matters is your shape against the industry you compete in.

    Knowing how you differ from your industry either through deliberate action or through inefficiencies is critical for competition. Techniques that work in other industries will not necessarily work in yours.

  • New patterns of structure that consider lifecycle are emerging.

  • Managing lifecycle is non-trivial.

    The use of lifecycle concepts enforces the treatment of an organization as a complex adaptive structure rather than a linear machine, but this in turn creates management complexity due to Ashby's law of requisite variety.

{{ bp-break }}
{{ Page[0].addThisButton }}


Organizational economics: Cloud platform


The platform is a critical area for competition

Combining the various threads within this paper highlights the critical importance of the cloud platform layer in the competition between organizations. To demonstrate this, a number of points need to be highlighted.

Whilst any innovation is a potential source of future worth, it is highly uncertain. As that activity spreads, it becomes more certain but with a declining differential value. Hence there exists an inverse correlation between future value and certainty. The correlation between certainty and future value and the necessity for innovation in order to compete tomorrow (creative destruction) means all organizations are forced to gamble.

Commoditization increases the rate of innovation but also tightly couples expenditure to use (utility charging). Hence the cost of any innovation gamble can be reduced by building upon a commodity provided through utility services.

As adoption is a means of identifying success, the larger the ecosystem surrounding an organization's platform then the more capable that organization is of identifying the next wave of successful innovation.

A company could exploit the ILC model to avoid the yo-yo between innovation and efficiency by first developing a platform of core utility services. It is critical that:

  • The methodology used to provide core services is not imposed on the ecosystem, hence reducing innovation.

  • A widespread ecosystem is encouraged, including competitors' use.

  • The ecosystem is nurtured through identifi cation of new patterns (innovations) and their subsequent provision as core services. Management of this transitional phase is key.

  • Where possible, commodity activities are provisioned through a competitive market and innovation pushed out to the ecosystem.

{{ bp-break }}
{{ Page[0].addThisButton }}


Organizational economics: Cloud playbook


Combining both the elements of organizational economics and the exploration of cloud computing, an initial playbook can be provided for how to deal with the cloud.

Realize that activities are constantly evolving. Any strategy implemented must be cognizant of this to avoid creating a future 'legacy'. Consideration should be given to whether the structure and techniques that are in place are capable of dealing with a constant cycle from innovation to commodity. Ideally, all activities should be identified into their relevant lifecycle using a process such as that described here.

Identify those activities that are suitable for utility provision. Look for ubiquitous activities within your industry that provide little differential advantage and should be suitable for commodity provision. Existing BPM and SOA efforts can help in this process of identification. Care should be taken to treat IT as a mass of activities (that is, some will be suitable for cloud provision, others will not) and to unbundle activities where possible.

Categorize those activities into infrastructure, platform, software and process,and highlight examples where consolidation is possible.

For identified activities, initially target those with highly variable demand.

Examine both enterprise and consumer markets for potential utility service providers. Consider the use of a broker and whether your organization wishes to deal directly with suppliers.

Consider also whether becoming a provider in this space makes strategic sense and whether these activities could become a viable platform – for example, cloud computing was not core to an online bookseller.

When examining a provider, give weighting to:

  • Provision of a standardized component (avoid customization).

  • Commodity provision (avoid oversubscription).

  • Existence of a market of providers (pricing competition, second sourcing).

  • Use of de facto standards (avoidance of later transition costs).

  • Options for providing equivalent services internally (hybrid environment).

  • Issues of strategic control (open source is preferable).

Before choosing a provider, consider transitional and outsourcing risks. Examine what orchestration tools are needed to manage this new environment, whether a hybrid environment will be needed, what governance procedures are required (for example, data protection), the mechanics of switching, and whether your organization has the capabilities to manage the Capex to Opex conversion.

Give consideration to the purpose of each new service. The rules for the provision of a commodity service (for example, an internally provided service would make extensive use of ITIL, SLAs, metrics, Six Sigma etc.) should differ from the rules governing the use of the service to create new innovative activities (for example, agile development, self-service).

Start small.

{{ bp-break }}
{{ Page[0].addThisButton }}


Organizational economics: Future


The cloud is just one of many waves of change

There are many approaching waves of change, some near term and some long term. For example, in the near term we have access to large datasets and in the long term changes to the means of manufacture.

The ability to process large sets of data has been a source of significant advantage to companies (for example, Google and Yahoo!). We have already seen Google extend its reach through these capabilities into providing health information, inflation data, house pricing and mapping. This trend of open access to complex datasets through meaningful services will continue to spread, whether in biochemical testing or market analysis.

Currently, manufacturing can be best described as using purpose-built facilities with techniques that are not widely understood. Additive fabrication (the process of printing solid objects, often commonly called 3d printing) is a technique behind a rapidly growing $1 billion49 industry that has also bled over into the consumer ecosystem50. By converting the process of manufacturing to simply pressing a button on a printer, this wave of commoditization threatens to reduce barriers to entry and increase competition in many industries. new forms of hybrid printers which combine both the printing of objects with electronics are under development and likely to emerge in the near future. The most extreme form of this is the RepRap51 project, which has successfully created a form of self-replicating printer, suggesting a future where the means of manufacture becomes viral.

Access to large datasets (a near-term effect – under three years) and commoditization of the means of manufacture (a long-term effect – over ten years) are two examples of the future impact of lifecycle. This process of evolution is continuous and whilst the cloud may well be the fl avour of the day, it is simply part of a larger cycle.

  • 49. Wohlers Associates, Additive Manufacturing State of Industry Report, 2010
  • 50. uPrint, Dimension Printer, 2009
  • 51. Dr. A. Bowyer, RepRap project
{{ bp-break }}
{{ Page[0].addThisButton }}


Organizational economics: Research


Key questions

This paper and our ongoing research within this domain raise many questions (see list above) about modern management practices. During the next phase of research we intend to conduct a wide range of in-depth interviews in order to help identify emerging patterns and provide guidance to our clients.

The concepts described within this paper have already been used in industry, and whilst the new patterns of organization are emergent (that is, common repeating examples can be found), it is questionable whether the patterns have been deliberately implemented or are simply the result of accident and organic growth. Furthermore, whilst elements of the ILC pattern can be seen in some industries (for example, in the existence of technical research groups and shared services), there is often a missing section (the settlers) which critically manages the transition from innovation to commodity.

It should also be noted that there is a lack of management books on how to manage the transition from innovation to commodity; there are no known playbooks on what strategies can be deployed in each stage of lifecycle to maximize advantage, nor are there any guides on how to implement the new organizational patterns. As with any activity, as knowledge and application of these concepts spread, their differential value diminishes. By the time management books, guides and best practices are written, there is often little differential value left. In the words of Oscar Wilde, "education is an admirable thing, but nothing that is worth knowing can be taught".

This suggests that these concepts currently represent a possible source of competitive advantage. LEF will be researching each of these areas and we encourage clients to distribute this paper to relevant constituencies within their own firms, to enable more voices to become part of the discussion.


Simon Wardley
Simon Wardley, based in the UK, is a Researcher for Leading Edge Forum and the lead practitioner for Wardley Maps advisory service helping clients anticipate market and ecosystem developments. Simon’s focus is on the intersection of IT strategy and new technologies, and he is the author of multiple reports including Clash of the Titans – Will China Dethrone Silicon Valley?  where he assesses the hi-tech challenge from China and what this means to the future of global technology industry competition. His previous research covers topics including Of Wonders and Disruption,  The Future is More Predictable Than You Think - A Workbook for Value Chain Mapping, Beware of Geeks Bearing Gifts: Strategies for an Increasingly Open Economy, Learning from Web 2.0 and A Lifecycle Approach to Cloud Computing. Simon is a seasoned executive who has spent the last 15 years defining future IT strategies for companies in the FMCG, Retail and IT industries.  From Canon’s early leadership in the cloud computing space in 2005 to Ubuntu’s recent dominance as the #1 Cloud operating system. As a geneticist with a love of mathematics and a fascination in economics, Simon has always found himself dealing with complex systems, whether it’s in behavioural patterns, environmental risks of chemical pollution, developing novel computer systems or managing companies.  He is a passionate advocate and researcher in the fields of open source, commoditization, innovation, organizational structure and cybernetics. Simon is a regular presenter at conferences worldwide, and has been voted as one of the UK's top 50 most influential people in IT in recent Computer Weekly polls.