Cloud computing—a utility model for computing capacity, software and business functionality—is now a pervasive presence in popular business literature and is being increasingly used by organizations worldwide to implement IT solutions. To date, however, there has been little substantive and objective research into the business implications of cloud capabilities to help executives make the right choices among technologies and services to meet specific business objectives.
Our latest London School of Economics/Accenture research—based on a survey of 1,035 business and IT executives, as well as in-depth interviews with more than 35 service providers and other stakeholders—supports the position that cloud is finally delivering on much of its promise. The business community appears more prepared to put cloud capabilities to work due to changes in internal and external factors—competition, innovation, globalization, user demographics, management readiness and supplier ecosystems.
However, a degree of hype remains, with the consequence that some are conflating potential long-term and short-term benefits in unrealistic ways. Among our survey respondents and interviewees, IT executives invariably warned us against assuming that full cloud functionality would be delivered quickly. Business executives, on the other hand, understandably wanted to see benefits from cloud realized within much shorter time horizons. This conflict between expectations and reality is a familiar story.
Many management, organizational and operational implications must be accommodated if cloud computing is to fully realize its value. Organizations and their CIOs need to plan carefully over the next three to five years to build phased plans that presume a mix of short-term and longer-term business impacts and benefits.
What kind of transformation?
Many executives and industry observers refer to cloud computing as a “transformational” technology. But what kind of transformation does cloud represent and how can cloud’s capabilities be harnessed most effectively?
One type of answer to those questions was articulated well by one of our interviewees, Tim Barker of salesforce.com: “Cloud computing in its best form lowers the barrier to actually getting the business what they want.” That is, through the use of cloud computing, the IT department can become a stronger force for delivery of tangible business benefits.
This view was supported by our survey findings: more than 50 percent of business respondents believed that cloud would enable them to focus on transforming their business and not just their IT.” IT executives largely concurred: almost half recognized that cloud could enable far faster implementation of business applications.
Two value streams
Our research supports the view that cloud computing is the maturation and convergence of two distinct value streams. The first relates to the technological infrastructure. Over the past 10 years, several technology developments have made it possible to externalize computing and storage capabilities “into the cloud,” where companies gain economies of scale in terms of IT support, energy consumption and speed.
By itself, however, computing infrastructure capability does not deliver the full range of benefits that cloud computing can provide. For that to happen, a second value stream needs to mature as well: a service perspective on computing. This focuses on the delivery of computing capability as a service that is consumed as and when required, instead of being provided as a one-size-must–fit–all capability.
The unique proposition of cloud computing is realized when both streams are relatively mature. That is, the service perspective of offering computing resources as and when they are needed is coupled with technologies that enable the variable use of virtualized servers over the Internet.
To see how these two streams interact to provide distinctive value, consider the case of media agency RAPP. The company uses the virtualization capabilities of cloud computing to address the variable processing demands associated with providing video streaming services for movie launches. If a movie launch is particularly successful, with many thousands of people wanting to stream the movie trailer, the company can scale its operations rapidly; if the movie is not as successful at the box office, then the service element of cloud computing means that they are not paying for unused infrastructure. RAPP can purchase as many, or as few, cloud services as it requires.
Four key benefits
Cloud providers are offering several types of services, with varying levels of embedded capabilities. At the top of the stack is software as a service (SaaS), in which the cloud provider runs all elements of a particular service (as provided by a company such as salesforce.com), with the user interacting with applications through a simplified, browser-based interface. The next layer of offering is platform as a service (PaaS), where a particular computing and application development platform is offered by the provider but managed by the user.
A third type of offering is infrastructure as a service (IaaS) which uses virtualization technologies to simulate on-premises hardware. Finally, a hosted service simply runs servers for a customer in an outsourced data center.
For all these service types, benefits are delivered and can be measured across four principal criteria:
- Equivalence: services that are at least equivalent in quality to those a customer would experience by a service running locally on a PC or server.
- Abstraction: organizations want simplicity—they want to have hidden from them the unnecessary complexity of the lower levels of the application stack.
- Automation: the desire to have a service automatically managed.
- Tailoring: customers want a service that is specific to their enterprise and its needs.
So where is cloud going and what innovations are possible to enterprises by choosing the right kind of service and managing it well? Here are a few insights into where our research suggests cloud technologies and services are headed.
When equivalence is achieved between an organization’s local data center and a cloud provider’s offerings, it is possible to create enterprise services by “mashing up” the services from a variety of cloud providers to create what has been termed a “cloud ecosystem.” This integration allows the tailoring of services to specific business needs using a mixture of SaaS, PaaS and IaaS.
Although this kind of cloud mash-up can create much more agile organizations, considerable technical skill is required from both a process and technology perspective to integrate such services. Further, the resultant ecosystem is limited by the service quality of the weakest component. Such ecosystems have been termed “Business Process as a Service” (BPaaS), reflecting the focus on business-specific services.
As equivalence is achieved between existing internal data centers and cloud-provided services, it becomes possible to use a hybrid mix of internal machines and cloud-provided machines within a business process. This so-called “hybrid cloud” allows parts of the process to be run internally (for example, handling sensitive data), with other parts of the process being run at low cost externally. With this kind of arrangement, cloud bursting is possible, where services are run internally but then jump or “burst” to the cloud when internal capacity is insufficient to meet demand.
For many enterprises that already operate large-scale data centers, the economies of scale delivered by IaaS and PaaS are limited (particularly if a company already operates multi-tenanted servers), whereas the risks of outsourcing to the cloud are perceived to be high. In response, some vendors have created software to operate existing enterprise data centers as though they were a pay-as-you-go cloud provider.
Our research highlights the need for companies to start with a set of more limited objectives and then grow into the cloud. This kind of incremental approach can promote learning, mitigate risks and reduce the difficulties of replacing existing technologies, organizational culture, structures and sourcing arrangements.
It’s also important to get detailed information about suppliers’ capabilities and consider these in the context of a company’s own business strategy and technology trajectory. Insist on a very strong focus on service from the supplier, backed up by contracts and service-level agreements.
Early in the process, involving business sponsors who have compelling business goals for cloud can help drive adoption and realize the business potential residing in cloud technologies.
Above all, because cloud is a service-based phenomenon, success will ultimately depend on everyone—clients and providers alike—adopting a strong service ethos.
About the authors
Professor Leslie Willcocks, Dr. Will Venters and Dr. Edgar A. Whitley are in the Outsourcing Unit of the Department of Management at the London School of Economics and Political Science.