Price Tag the Cloud

A key characteristic of cloud computing is the potential ability to price, deliver, and consume IT resources in a flexible, usage based manner. Rather than paying for an infrastructure, customers pay only for the needed computing power - use more, pay more and use less, pay less.

The implications of this shift are nothing short of transformational. Consumption based pricing enables transparency and encourages demand management: if the business understands the cost implications of the way IT is used, better choices will presumably follow. In addition, usage based pricing theoretically eliminates the cost of insufficient or idle capacity, since resources are allocated only as needed.

The problem is, we're not quite there yet. Accurate and consistent pricing models for cloud services remain elusive, and significant organizational obstacles must be overcome to leverage the potential of flexible pricing.

Customers exploring cloud initiatives need to understand market dynamics as well as the challenges involved in comparing various offerings and assessing the true total cost of a cloud solution.

Pricing Models

Today's marketplace is characterized by myriad pricing models with a wide range of different structures. The diversity is based largely on whether the cloud offering is public, private, or a hybrid. Provider A may charge by processors reserved, while provider B charges by the CPU hour. Examples of new total cost of ownership (TCO) calculators are cited below:

• CloudSizer.com (IDEAS International)
• Microsoft Azure Pricing Calculator
• GoGrid
• Amazon Elastic Compute Cloud (Amazon EC2)

Microsoft recently released its Windows Azure Pricing Calculator, adding to the growing list of simple, web based pricing calculators for cloud platforms. Amazon Web Services, a public cloud pioneer, has had a pricing calculator in place since 2009.

Of course, in true cloud fashion, it's still in beta. Cloud-hosting providers have also embraced the 'transparency' trend, with Rackspace, GoGrid, and NaviSite all displaying public pricing on their websites. Some have had this level of cost visibility in place for years; others are just now going public with their cloud pricing.

These pricing tools are providing unprecedented visibility into specific components of pricing and the cost required to run key parts of the infrastructure and applications in the cloud.

That said, the ability to gauge utilization accurately enough to bill for it in a true utility manner remains a challenge. Most internal and external IT groups don't have the monitoring tools to accurately measure and log CPU minutes. Moreover, a CPU minute-on-one-machine can represent more or less processing power than a CPU minute on another machine, so ensuring an apple-to-apple comparison of different service offerings is problematic.

It's also important to note that these calculators only provide certain components of the pricing and don't represent the real total cost to operate in the cloud.

Allocation

Some current cloud models use allocation, such as a server 'instance' or a compute 'slice', as the basis for pricing. Here, the resource that a customer is billed for has to be allocated first, thus allowing for predictability and pre-approval of the expenditure.

However, the term 'instance' can be defined in different ways. If the instance is simply a chunk of processing time on a server equal to 750 hours, that equates to a full month. If the 'size' of the instance is linked to a specific hardware configuration, the billing appears to be based on hours of processing, but in fact reflects access to a specific server configuration for a month. As such, the pricing structure doesn't differ significantly from the traditional server hosting.

Terminology is also an issue. Since no clear standards exist on how cloud services are defined, the same set of words one cloud vendor uses to describe a truly innovative, cloud based delivery approach can be used by another vendor to hype a traditional delivery mechanism. In other words, customers need to ask: Are we truly leveraging the potential benefits of cloud computing? Or are we simply getting traditional services with a cloud label attached?

Despite these issues, cloud services can and do offer increased flexibility and a more streamlined process for growing an IT infrastructure, as the provisioning of new servers can often be done in a matter of hours or days, instead of weeks or months, even if the services are billed via an allocation model. In addition, the move toward cloud services has the practical impact of driving standardization across multiple customer environments and making services more utility-like. This allows vendors to leverage greater economies of scale and deliver significant savings to customers.

Cloud TCO

In today's buyer-beware cloud environment, customers need to understand the details of each pricing scheme in a comparative context, along with the terms of the agreement, to determine exactly what they're getting (or not) for their investment. Otherwise, the cost of the cloud solution risks becoming in addition to, rather than instead of, the existing environment.

The process is analogous to buying a car online. The base price looks extremely attractive, but when all the options, tax, title, license, maintenance, and fuel consumption are added to the base price, the cost may be as much as 50% more.

Similarly, when pricing a true enterprise cloud solution, the total costs involved have to be considered. These include training staff on re-engineering applications for the cloud, re-working the authentication framework, integration with legacy applications, and building in business continuity and disaster recovery solutions.

The cloud pricing calculators described earlier don't adequately account for these incremental costs. Examples of costs frequently not calculated in cloud solutions include:

• Internal Level 1 service desk and technical resources to interface with the provider's Level 2 desk
• Migration costs to the cloud
• Additional network bandwidth (data transfer rates in and out)
• Remaining book value of stranded assets and software
• Remaining amortization of applications
• Additional staff to manage new contracts
• Cost of fulfilling minimum volume commitments in current contracts

A decision to process and store data in the cloud does not eliminate other costs inherent in the IT environment. While cloud calculators provide detail and transparency into the base services price, that's not enough. CIOs need to understand the true total cost of ownership of a cloud solution - including the cost to transition and the cost to run operations in the cloud.

It's also imperative to accurately assess and evaluate cloud options and benefits in an apple-to-apple context. And, the key questions include: Which pricing model is appropriate? How does offering A truly compare to offering B? How will choices support or undermine the overall IT and business strategy?

Looking Ahead

Organizations that will drive usage-based pricing to further maturity will be those that are willing to give up predictability for overall savings, and are willing to work with cloud providers that are hungry for market share. Such pioneers - Amazon, for example - have businesses with revenue-generating applications that need to scale rapidly. For them, scalability trumps predictability, and growing revenues can cover higher costs.

In the meantime, for customers who are skittish about unanticipated spikes in expenditures, service providers are offering pricing models that set off alerts when a certain level of consumption is likely to be reached.

The utility model will be most attractive for revenue-generating applications that need to scale quickly, development environments that need to scale up and down, and compute-intensive applications that can run processes in parallel across multiple machines (today's grid applications). For more common types of production processing, mechanisms that allow clients to retain an effective control over increases in demand will be needed.

From the service provider perspective, true usage based pricing will be driven by vendors aggressively seeking market share. These players will push for increased flexibility coupled with decreasing minimal commitments. Vendors defending market share, meanwhile, will be pulled reluctantly down this path.

SCOTT FEULESS AND

STANTON JONES

The authors are principal consultant,

Compass Management Consulting and analyst,

emerging technology, ISG, respectively

maildqindia@cybermedia.co.in