Setting the Right Expectations for Cost Reductions from Cloud
Dave Brown and Stan Lepeak
Shared Services and Outsourcing Advisory Group, KPMG
Potential benefits from cloud computing are a little like the real estate market and credit card offers. There are good savings to be had, but not all offers are what they appear to be, and it is critical to understand what you’re really buying.
The benefits from cloud are not always as advertized. Indeed, many providers and pundits tout that “the cloud” is not only revolutionizing the way information technology is delivered, but its usage can promise savings of up to 70 percent over traditional IT models while also unleashing a new era of IT flexibility and scalability.
In truth, the cloud in many cases is simply new marketing of existing and evolving information technology products and services. For example, service providers have long used the Internet to provision infrastructure and other IT capabilities as service offerings. What is changing today is the breadth and scope of available cloud based applications and services. When it comes to savings, cloud certainly offers opportunities to reduce costs, but if buyers expect to get automatic savings of 70 percent across the board, then their thinking is, well, a little cloudy.
Ultimately, the savings potential varies depending on the cloud model deployed (e.g., public vs. private), pricing structure utilized, usage levels, and other key factors. Following are some important considerations for different models.
Software as a Service (SaaS)
Utilizing SaaS, organizations can streamline the process of deploying and managing point and enterprise software solutions. Efforts to install, customize and maintain the systems are lessened and simplified. Organizations can move towards a model where they can access the applications “on-demand” through an Internet browser, which can generate potential savings of 30 to 40 percent over traditional packaged software deployment models.
SaaS applications are not new. Salesforce.com and Google mail have been around for years, and SaaS applications to support human resource activities, for example, have been gaining market traction as well. What is changing is the breadth of applications available, their greater maturity to support corporate versus just consumer usage, and the accessibility of these applications thanks to devices like mobile phones and tablets. Additionally, buyers are starting to become more comfortable with security and data privacy requirements that prohibited them in the past from entertaining such delivery models. While these are all positive trends for the SaaS market and its users, they do not necessarily imply the inherent achievement of major cost savings versus traditional software models.
Looking ahead: There are additional pressures working against the potential for significant cost savings from utilizing SaaS applications. As SaaS applications become more pervasive, the real costs of integrating them into legacy enterprise systems environments increase. There are also direct costs related to managing multiple SaaS providers providing piecemeal solutions alongside of legacy software solutions and often outsourced application and business services. In addition, there is the ongoing pressure and allure to customize SaaS applications to support custom business processes. This is nothing new. One of the touted benefits of ERP systems dating back to the 1990s was their more standardized process models, though the reality was that many firms heavily customized their ERP systems.
SaaS pricing structures will also evolve in the next three to five years. Today, for example, the most successful SaaS pricing structures are based on usage – either the number of users or the number of times an application is accessed. Sometimes, however, what’s offered as a usage agreement may actually be a software license agreement in disguise. Going forward, pricing will change as new players enter the arena and experiment with creative pricing structures.
Infrastructure as a Service (IaaS)
Some of the potentially highest cost savings from the cloud can come from IaaS. In this model, instead of buying and maintaining internal IT infrastructure hardware and software, an organization can buy infrastructure services such as storage capacity from a third-party provider. This in itself creates cost savings opportunities, especially if a client is willing to operate in a shared or public (vs. private) cloud operating model.
One touted benefit of IaaS is the speed of development and testing of new applications. To test a new application, for example, an organization might traditionally take six to eight weeks to procure and set up a server for a couple of months of testing – and then spend more time tearing it down. With IaaS, on the other hand, an organization can (theoretically) log into a service provider’s test environment in 24 hours or less. And because the servers are spread across multiple clients, the provider isn’t setting up a new server every time a company has a test or wants to start developing. The result can be up to 50 to 60 percent reduction in cost to the buyer – as well as faster speed to capability.
However, the real world may prove different. The technical aspect of raising and deploying a new server is one thing; the contractual and pricing aspects of doing that across a large organization is a different and more complex challenge. Here again, the potential of cloud savings meets the reality of utilizing the cloud in a large operational organization with policies, procedures and procurement rules and regulations.
But perhaps the biggest hurdle to reaching the 50-60 percent savings range is that many larger organizations are uncomfortable with the public cloud model, especially when sensitive data is involved. Many organizations only want their data and systems in the cloud if it’s in a closed and dedicated environment (private cloud). However, usage of private clouds typically generate less profits for service providers than a shared public cloud environment. With private clouds, service providers have to invest in dedicated equipment that keeps data separate, which reduces the savings to 20 to 30 percent over traditional infrastructure models.
The public versus private debate leads to a variety of additional questions around cloud. What kind of data does an organization want to put in the cloud? What are the rules for privacy and data security? Organizations must answers these types of questions when assessing cloud opportunities and potential savings.
Looking ahead: The IaaS pricing structure enables organizations to buy capacity, but involves many additional complexities. What if, for example, a company requires the use of a certain platform such as UNIX or “WinTel,” or wants a specialized type of storage? At what point does usage trigger a volume discount? Because of these additional factors, assessing actual cost savings remains challenging, and it will take time for these pricing structures to mature as the industry gains experience with their usage.
Another word of caution on IaaS pricing: The do-it-yourself nature of cloud infrastructure may create overspending that depletes any savings. For example, if someone in a business unit wants to set up a test environment without going through IT or centralized procurement, they can visit a service provider’s website, identify their requirements, and pay for it with a credit card. However, the 40 hours they purchase may actually be twice what they need. Unchecked, this kind of practice can create pockets of IT overspending throughout an enterprise and highlights the importance of strong consumption management practices when adopting any cloud service model.
The cloud clearly creates opportunities for potentially significant cost savings over traditional IT infrastructure and enterprise software models, but savings will vary significantly across different deployment scenarios. The highest savings will go to those organizations that are willing to adopt public cloud models, often not an option for many large established organizations, as well as those who can live with more standardized and less customized computing environments. Regardless of the cloud model deployed, users should stay abreast of new pricing developments, as the savings offered today may be very different tomorrow.
Dave Brown is a principal in KPMG LLP’s Shared Services and Outsourcing Advisory network and can be reached at firstname.lastname@example.org. Stan Lepeak is research director in KPMG LLP’s Shared Services and Outsourcing Advisory network and can be reached at email@example.com.