Clearing the Fog from Cloud Computing
Analysts tout it, vendors promote it, the media report on it. But discussions about cloud computing are often obscured by IT jargon. Clarity reveals the cloud's silver lining.
There is no question that cloud computing has made a name for itself. Companies worldwide are turning to public and private clouds for their IT needs. In fact, 80 percent of Fortune 1000 companies are expected to be using cloud-computing services as early as 2013, and 20 percent of them will do so without ever owning a single piece of hardware. By 2014, the value of public cloud services is expected to grow to $55.5 billion from $16.5 billion in 2009. That's an annual growth rate of 27 percent—about five times the growth rate of IT services.1 The same type of technology resides within most companies. Often called a private cloud, it too is growing fast, influenced by the growth of public clouds and the new capabilities cloud provides.
What is cloud computing? The most often cited definition comes from the U.S. National Institute of Standards and Technology: The cloud enables convenient, on-demand network access to a shared pool of computing resources—networks, servers, storage, applications, and services, among others.2 Several key features are crucial to the cloud offering and to today's dynamic business environment (see figure 1).
Cloud computing represents a significant and needed step toward the "utility-ization" of IT services across all layers of the architecture stack, meaning IT functions are becoming more standardized and services-based. Just as no programmer would expect to develop code for windowing, scroll bars, and button clicks, many service components in the cloud, such as credit-card verification and billing, are like Lego blocks: pre-built and ready to be snapped into applications.
Access to the cloud is fast and inexpensive. Cloud computing does not require large capital investments in infrastructure, technology, applications, or platforms. The cloud's architecture offers flexibility, which is essential for accommodating sudden shifts in demand, and is accessible from a broad range of devices. Companies in the cloud can respond faster and more effectively to business needs and customers, which by extension often means improved productivity, more innovation, and faster go-to-market strategies (see sidebar: The Cloud's Silver Lining).
So far, the cloud is primarily deployed in areas with minimal concerns, such as email, backup, storage, and testing. In the next few years, as standards are established around data security and storage, we expect to see the cloud in many other areas—and, not surprisingly, creating more profitable businesses.
Cloud Computing Business Models
The financial and capability benefits realized by companies leveraging cloud computing can be profound. There are few upfront capital requirements (at least for external cloud computing), and the costs are more variable, making cloud computing very attractive from a yearly budgetary cycle perspective. The total cost of ownership can also be increasingly beneficial as upgrades are borne by large scale architecture changes for the provider.4
In developing a cloud computing business model, we often demonstrate to our clients four options depending on their focus, ambitions, and maturity: Virtual Voyager, Architect, Network Native, and Mobile Mover (see figure 2). Each one comes with opportunities, implications, and tradeoffs, depending on the extent to which the organization wants to capitalize on the cloud and the consequences of extending, or even reinventing, an existing business model. The following offers more detail about each model.
Cloud computing, for all the hype, is a virtual computing resource that offers flexibility and speed via remote access to software, operating systems, processors, peripherals, and networks. The Virtual Voyager introduces five distinct platforms as building blocks (see figure 3):
Software as a service (SaaS). The most mature platform, SaaS consists of software packages hosted and delivered via a public or private network. Salesforce.com is a good example as it sells customer relationship management (CRM) software via the Internet on a per-use and per-transaction basis.5
Platform as a service (PaaS) bundles databases, web servers, and development tools and sells them as a package. Both Google App Engine and Amazon's EC2 are in the PaaS family, selling access to software but not the installation headaches.
Infrastructure as a service (IaaS). IaaS is all about virtual servers, extended networks, and remote storage. No more large hardware investments, maintenance, or infrastructure upgrades. For those who miss the hardware, no worries: Some IaaS providers offer supercomputing services for a per-hour charge. AT&T's Synaptic Hosting is an IaaS poster child.
Business process as a service (BPaaS). Like IaaS, BPaaS is a hardware-free, end-to-end service that has helped Avaya, Echopass, and Verizon set up call centers complete with phone lines and computing capacity in as little as 60 days and activate additional resources quickly when demand spikes.
iMobility. iMobility represents a dramatic shift in personal communications as smart phones, tablets, netbooks, and laptops allow people to stay on the go and connected. iMobility is possible because of cloud vendors with immense global IT infrastructures, including data centers and cell towers.
The Architect goes a step further to "virtualize" select business components into the cloud. It sidesteps technology silos to offer an external network platform from which to automate processes and manage business processes. When the data model and the workflow management application are both cloud based, workers across the organization can log in and work seamlessly on different aspects of a project.
Virtualization introduces a new layer of abstraction into computing that can make it more mobile and resilient. A task used to be done on a specific machine's processor and its data stored in a specific hard drive. With virtualization, the task and storage exist independently of particular hardware—they can be reassigned if capacity is a problem or if there is a hardware failure. End users no longer need to worry about where data is stored and processed, but they will of course want to know that their systems will be available and that redundancy and backup are provided.
For example, a bank managing workflow among its many retail branches might use the cloud to speed up processing. Architects and engineers designing a production plant can access the same data and computer-aided design packages to achieve a faster, more collaborative outcome no matter where they are in the world. Human resources, finance, claims management, research, procurement, and even sales and marketing functions can all benefit from a cloud-based setup. And the service can be achieved relatively quickly because it resides outside the existing IT landscape.
The cloud is also an excellent venue for capturing customer insights—providing an integrated view of the various channels customers use for researching, commenting on, ordering, and rating products. A visit to an e-commerce website generates a treasure trove of data from per-page dwell time, link patterns, and products viewed to items added to wish lists and shopping carts.
The Network Native is a cloud-based business model with very few processes occurring outside of the cloud. Network Natives are focused on one-to-one interactions with their customers and the superior, timely integration of information.
For Network Natives, cloud computing means conducting processes externally, with the potential to open up new business opportunities. For example, 3M launched a cloud-enabled service called Visual Attention Service (VAS), which gauges customer reaction to visual imagery. Outside advertisers and designers visit the 3M website and, for a fee, can upload an image to be analyzed. In just minutes, the system provides a detailed report, heatmap and all, of the design, highlighting where the average person's eye is drawn. Because the service is cloud based, 3M's upfront investment is minimal and its pricing model is pay-as-you-go.
Another example comes from the healthcare industry and its many players, from doctors and pharmaceutical companies to insurers, hospitals, and patients. With patients' medical records scattered among various players in multiple locations, cloud providers are reaching out to healthcare professionals, offering help in organizing, storing, and sharing health information online. Quest Diagnostics, for example, posts lab results in the cloud, making the information readily available to both patients and healthcare professionals.
Pharma companies, especially those heavily invested in R&D, are using the cloud and team collaboration tools to shorten the time to market for drugs. In mobile telecom, Giffgaff, a UK mobile operator, operates totally through the cloud and social media—no retail distribution, no big-budget advertising, no sprawling call centers. About 20 percent of Giffgaff customers, known as giffgaffers, get all the information they need from a user-generated pool of knowledge in their online, cloud-powered community.
For smaller companies, the cloud can help them compete on a level playing field and break down barriers to entry, allowing them to rent capabilities via the cloud versus making large capital investments that can delay market entry.
The Mobile Mover is an extension of the Network Native, using mobility to create new business models. The main difference is that these customers are mobile while Network Native customers are both static and mobile.
As more people become digital consumers, the way products and services are delivered and consumed is changing—and so is how organizations set up their business models and supply chains. For example, iMobility—linking cloud-based business models to mobile consumers— requires both an always-on device and access to an always-available network (fixed line, 2G, 3G, 4G, or WiFi). For example, a customer who gets bad service at a restaurant might forget about or delay writing an online review until he gets home to his computer. But a customer with a smartphone can access a site such as Yelp or Zagat and instantly submit a review while waiting for a long-overdue check. Facebook, with its focus on collaboration and mobility, has been a leader in this area for some time and is now a platform for other applications, such as Animoto and Real, which take advantage of on-site interactions.
The co-creation of customers and value, often tied to real-time updating, is having a dramatic impact on customer expectations and company value chains.
Getting Up on the Cloud
At first glance, implementing cloud computing may seem daunting, but it doesn't have to be. In our work with clients, we position the process as it takes place in three stages:
Align the cloud-computing strategy with the business strategy
A good cloud-computing strategy does not ignore the business strategy—it embraces it. There is a huge financial advantage in leasing capacity and storage from a shared environment. Banks involved in algorithmic trading need real-time market data, so they align their cloud strategy with their business strategy. Online software companies use cloud computing to emphasize performance and scale quickly. These firms launch many different products hoping one catches on, and when one does, they have to quickly scale to gain the advantage. Instagram, the online photo sharing software, is a good example because it used the cloud to capture its window of opportunity. Companies that offer mobile information services, such as driving directions, use cloud computing to build global capacity. Retailers, especially low-cost leaders, capitalize on the cost reduction and cost avoidance aspects of cloud computing to expand their reputations and customer base.
Identify cloud candidates
First-time cloud strategists should begin with quick-win projects. Replacing servers that have low usage levels or are near or at end of life with an IaaS offering is a good starting point. Migrating software development or testing to the cloud is also a good beginning strategy. These areas are often built up, taken down, and rebuilt, which means they offer three advantages: flexibility to use on-demand resources, automated processes to reduce the setup time, and reduced time and expense. There is no need to purchase, rack, or connect servers because these are not needed after development and testing.
Medium-term cloud projects will require more time and effort. Applications already running on a virtualized server are good candidates because end users and system administrators are familiar with virtualization and applications and the systems are using standard protocols. Longer-term cloud projects will require significant investment, but they also deliver significant returns. These are the cloud-based services that are available 24-7 in all locations, such as music players. They require a global infrastructure and advanced knowledge of cloud-computing technology.
Choose a cloud business model
Select the model that makes the most sense for your company, customers, and IT department (see figure 4). Ideally, the corporate IT organization will lead cloud-service design, implementation, operations, and process management. This way, it is possible to consolidate scale, buying power, skills, and internal experience to deliver the best possible service to internal customers—the business units. Corporate IT might also consider funding cloud venture projects in collaboration with business units. This reduces the barriers to adoption that can arise with already tight IT budgets. And to increase the likelihood that the business units get the most out of cloud computing, projects should be staffed with enterprise architects who understand the business needs and the drivers behind cloud service designs, including architecture, security, and compliance requirements.
Lessons to Be Learned
As with every move to a new system, technology, or process, the experiences of those who have already been down this path cannot be overstated. In the case of cloud computing, we have several lessons to share from our work with clients:
Management style. Cloud computing changes everything for corporate IT, which means policies and management style will have to change. All new initiatives and technology strategies must be viewed through a cloud lens as cloud computing becomes a viable option. Regulatory and corporate compliance rules will have to reflect the differences between cloud computing and legacy systems. For example, we helped one client develop data-classification policies to indicate which types of data can reside in a multi-tenant environment for security and access reasons. At the same time, requiring physical inspection of data-center facilities may no longer be reasonable as cloud vendors often have data centers across the globe.
Operating a cloud-computing environment will also require different capabilities. Manual processes will be automated, and corporate IT professionals will need very good vendor management skills to manage multiple virtual components from different cloud vendors. It may also be necessary to rethink traditional service-level agreements (SLAs) because cloud-service providers could be held accountable for the end-user experience.
Partnerships. Corporate IT and the business must become close partners. Cloud computing can transform a business, but only if the business and corporate IT are in lockstep to capture all the benefits and synergies. Make no mistake, business users will become more important to corporate IT as they deal directly with technology services. And costs will rise if IT is not involved in the selection, acquisition, and support of cloud services. The proliferation of applications must be avoided at all costs.
Flexibility. IT will have to meet users' needs—even as customers are in more locations, asking for more access and more information, and are always connected via their mobile devices. If a company's solution is insufficient in any way, customers will become dissatisfied, and sales will be lost. The key is to be systematic in developing data that identifies new usage patterns, business models, and revenue opportunities. Top companies ensure that the business and its IT partners work together to identify and respond to users' changing expectations.
Knowledge capital. As business processes are outsourced to cloud providers, IT will have to stay current—able to execute all business processes even as they are altered. Relinquishing knowledge and aptitude is dangerous. It is never pleasant to be held hostage by a low-performing cloud service provider.
Data. Managing customer data is essential. Although it is necessary to trust cloud vendors with company data, sound data-security policies must be in place to support acceptable risk levels. For example, customer data can remain anonymous before it is transmitted to the cloud, thus reducing the risk that data loss directly affects customers.
Moving Into the Mainstream
The cloud represents a significant shift in the way business manages computing and is quickly developing a mainstream presence. In building a cloud capability, there are business models to consider and steps to take, each with immediate opportunities and long-term implications. Although the cloud is not without significant challenges, it is also not without substantial business potential. What's important now is moving the cloud to the forefront of IT strategy planning. This is not an area where it is possible to catch up. In the cloud, once behind could mean always behind.
The authors wish to thank Jim Lee for his valuable contributions to this paper.
1 "Document Management in the Cloud and the Future of Cloud Computing," Cloud Issues, May 2012
2 The NIST Definition of Cloud Computing: Recommendations of the National Institute of Standards and Technology, Special Publication 800-145, U.S. Department of Commerce
3 "Royal Mail Virtualizes Server Farm to Save £1.8 Million in Four Years," Microsoft Case Studies, November 2010
4 See Finding What Every ICT Services Provider Craves: Healthy Profits.
5 See Saas: Managing the Anti-IT Application.