The cost of support and maintenance of a given platform typically represents a cradle-to-grave expense that can exceed the cost of the hardware by several times.
In today's world, businesses are paying more attention than ever to cost-cutting measures by controlling the cost of their assets. They must therefore strive to obtain systems that have enhanced reliability, faster deployment time, and lower cost of ownership.
Total Cost of Ownership (TCO) is a model that is used to identify and analyze the overall costs associated with ownership of Information Technology (IT) assets, from cradle to grave. Reducing the cost of ownership of IT investments will result in higher Return On Investment (ROI). Through TCO study, businesses can make better strategic decisions and control costs.
There are many factors that need to be considered during the course of a TCO study, and identifying those factors is the focus of this article. We will discuss each element contributing to TCO and explain why it should be part of the TCO study model for any system.
According to the Standish Group, enterprises collectively spend over one trillion dollars per year on IT. As competitive pressures mount within those industries most reliant on IT infrastructures, companies seeking to increase the value of their IT investments and to improve the efficiency of their server platforms have begun moving beyond straight-line comparisons of hardware costs to a more realistic examination of the full range of costs associated with operating a server. Analyses of this type reveal that the factors that influence the costs of operation are frequently much more complex than hardware cost alone, consisting of a combination of tangible and intangible factors. TCO comparisons therefore provide a real-world gauge to more accurately determine the anticipated expenses associated with specific server platforms.
Beyond initial hardware-acquisition costs, TCO studies typically focus on those factors that represent ongoing expenses during the life cycle of the server platform. For example, the cost of support and maintenance of a given platform typically represents a cradle-to-grave expense that can exceed the cost of the hardware by several times. Controlling support costs clearly provides an effective approach to controlling overall costs within an IT group. Other factors that bear careful examination when assessing the TCO for a server platform include operating costs, expenses associated with upgrading firmware, application-software costs, network-hardware expenses, and scalability costs. Cumulatively, these factors indicate that the actual hardware costs represent only one component of the aggregate expenses incurred by an organization following their selection of a server platform. By recognizing and addressing the costs associated with the remaining components, IT budgets can be more effectively managed, and considerable savings can be realized in a number of areas ahead of time.
Incorporating TCO considerations into evaluations of competing systems provides companies with a reliable means of assessing comparative costs in a number of areas, leading to the potential of considerable savings when provisioning and supporting their IT infrastructure.
Total cost of ownership for an enterprise application server includes cost factors that both directly and indirectly affect the bottom line. A realistic assessment of the actual expenses involved in purchasing, deploying, maintaining, and upgrading the hardware and software required for a server platform should encompass these factors:
- Purchase price of the hardware
- Maintenance and service contract costs
- Network infrastructure costs
- Fixed operating costs
- Deployment costs, including database conversions and application migration
- Integration expenses to existing systems
- Ongoing operating costs, as affected by reliability, availability, serviceability, and manageability of the platform
Deployment, initial configuration, and installation represent significant factors that directly affect the TCO value, both in the time invested in implementing an enterprise application and in terms of the specialized service costs to configure the system.
The streamlined development times associated with Web services also contribute substantially to reducing the time to market for large-scale applications. Applications can be deployed, launched, and tested on a limited-access Web server and, once they have been approved for release, installed on a target Web server for customer and business access. These reduced development times provide one more benefit worth consideration when making TCO comparisons.
Evaluating TCO Factors
Determining the total cost of ownership for an enterprise-capable server involves calculating both the tangible costs and the intangible costs of ownership. Tangible costs such as the following translate easily into hard figures:
- Purchase price of a server
- Cost of a service contract to maintain the system
- Network equipment required to support intranet or Internet access
- Fixed operating costs, which factor in elements such as power consumption of the hardware, cooling requirements of the system; and facility requirements, including rack mounts, space required for the system, and so on
In most cases, the tangible costs can be directly determined, based on invoices or estimates. These kinds of figures can be readily used to provide a base-line level of comparison between two competing systems.
Intangible factors in any TCO evaluation impact the bottom line in a variety of ways, but by their nature, they are more difficult to measure and interpret. Nonetheless, organizations can effectively address many of the intangible costs when evaluating comparable platforms, and by doing, so they can contribute to long-term savings. Intangible factors include the following:
- Conversion costs for database and application migration to a new platform
- Risk-mitigation costs to minimize service disruptions during platform deployments
- Integration costs in connecting a platform with legacy systems and existing company infrastructure
- Operation costs associated with the reliability, availability, serviceability, and manageability of the server platfor m
Tangible Costs of Ownership
Tangible costs of ownership provide the most accessible basis for comparing server platforms prior to launching a new initiative or deploying enterprise applications. As can be easily demonstrated, however, the purchase price alone serves as only an approximate starting point as to what the overall TCO value will be. Only after both the tangible and intangible costs are evaluated can the full cost structure and long-term value be effectively analyzed.
Purchase Price of Server Equipment
Evaluating different server platforms can often be a difficult exercise for a number of reasons. The range of standard and non-standard features can affect the price, as well as the cost of components required to carry out the essential tasks for which an application has been designed, including network-interface components, hard-disk storage, system memory, and so on. At the heart of any evaluation of this type, the cost-versus-performance ratio is a key consideration. Only by comparing systems of roughly equivalent performance can the actual value of the hardware components be accurately assessed. Inexpensive systems that lack adequate performance to handle mission-critical tasks clearly don't represent genuine value to an organization. High-performance systems that are capable of handling enterprise requirements but that are inordinately expensive also fare poorly in a TCO evaluation. Cost and performance cannot be separated in a rigorous TCO evaluation.
Service and Support Costs
Service and support costs can be examined in a number of different ways. For example, depending on the complexity of the installation, support costs can vary widely. Daily ongoing operations, usually covered in a service agreement, represent another aspect of the support costs that can be directly calculated and compared. Some system implementations have higher built-in costs than others. For example, systems that require supporting applications and upgrades to be installed manually on client workstations typically have relatively higher support costs than Web-based systems. Systems that utilize network-based deployment, Web services, or browser-based interfaces can manage upgrades through simple changes on the application server. The cost of documentation for a system–both the installation expense and the ongoing operation–is one more tangible cost.
In general, systems based on open standards and widely-available components fare better when considering service and support issues. Proprietary systems that use specialized hardware and software typically require service skills that are more expensive in the open marketplace. The same market forces and widespread competition that tend to lower service costs for commodity hardware and software make it more expensive to service and maintain proprietary systems.
Service practices and fees typically vary from company to company, making it difficult to equitably assess the relative value of services in a system-to-system comparison. One method that helps quantify service costs is to examine the service descriptions to determine what is included for a particular product. Many times, free services are included, which can add greatly to the value of the overall service offering.
Over the course of years, it has been proved that the cost of hardware represents about 15% of the total cost of an IT asset, while support costs are typically as high as 70% to 85% of the total cost.
Training and Routine Maintenance Costs
Some tangible costs require a bit more research and investigation to calculate accurately. For example, the cost of training personnel to perform routine tasks on a system and the expense associated with updating firmware contributes to the TCO. This is true whether the direct costs factor in the software expense or acquiring the necessary skills required to perform the upgrade. There are also hard costs associated with the operating system selected to run on the platform and the application licenses needed to support the corporate mission.
The selected database application with its licensing costs represents another software expense. For many projects, the software expenses can account for a sizeable percentage of the total investment. For specialized applications, software costs can sometimes increase to a level as high as the cost of the hardware system itself. While assessing database costs, some of the same factors that affect hardware cost versus performance come into play. The database performance must be equal to the task at hand, performing a given application with sufficiently fast response times.
Another very critical factor, particularly for enterprises operating in high-growth fields, is the cost of scalability. Systems that are designed to scale up and scale out inexpensively can benefit IT organizations during rapid growth periods, reducing the costs associated with adapting to dynamic change. Scaling and clustering represent two different approaches to managing massive increases in transaction volume, and each of these approaches has a different set of costs attached to it. To maximize the bottom line and to achieve optimal cost savings, the approach that delivers the appropriate combination of availability and fail-over strategy while increasing load-handling capacity should be the choice in most circumstances.
Other hardware costs to consider include the cost of integrating a server platform with an existing infrastructure and the overall interoperability cost. Seamless integration with existing legacy systems is often an essential requirement in IT environments, since many organizations have a substantial investment in these systems, and wholesale replacement is often not immediately feasible. Those system architectures that demonstrate benefits and advantages in terms of ease-of-integration can generate substantial cost savings. Interoperability also factors heavily in environments where a complex mix of hardware and software platforms must be universally supported.
Intangible Costs of Ownership
Intangible costs also have a very real impact on the bottom line, but by their nature they are often more difficult to calculate and to factor into a comparison. For this reason, they are sometimes referred to as "soft costs." The more effectively an organization deals with the range of intangible costs associated with a server platform, the more likely it is that extended cost savings and efficiency can be realized during the life of the server platform.
TCO factors that fall into intangible costs are described in detail below:
- Conversion costs
- Risk-mitigation costs
- Operation costs
Organizations contemplating a platform change face an initial conversion cost made up of the expenses involved in porting the required software and database content to the new platform. The actual costs associated with porting depend on the level of conversion required to retain as much of the earlier data and framework as possible. For example, moving a suite of Java* applications from one platform to another is less likely to incur significant conversion costs than adapting an Oracle* database for conversion to SQL Server*. Another key element of the initial conversion is the deployment cost-those expenses required to put the new platform into service-and the training costs associated with providing personnel with the skills to operate the new hardware and software. Training costs for teaching personnel to operate unfamiliar applications or to work in a server environment to which they are unaccustomed can accrue quickly.
During the conversion, additional tools may be required to complete the installation, to perform setup operations, and to initiate backup and restore procedures. These costs factor significantly into the total cost equation, as do the maintenance costs related to the upgrade and the expense associated with acquiring skilled personnel to assist in the conversion.
Risk Mitigation Costs
Launching a new server platform is not a trivial exercise, and the process of moving a substantial part of a business operation from one platform to another entails a certain degree of risk. Solutions that mitigate risk factors can influence the total cost of ownership, even though those risk factors may not translate into a hard set of figures. The selection of a solution that lowers risk factors is a sound strategic measure and a means of moderating potential disruptions that may arise from developing a new infrastructure for critical business operations.
Investments that are tied to a sole vendor represent an identifiable risk. Solutions that favor multiple sourcing lessen the risk, providing a clear alternative path in the event that a primary vendor drops product support, fails to deliver on agreements, or raises the costs associated with product upgrades or maintenance.
Another measure of investment protection is the ease by which an existing system can be upgraded to the next level of processor. This type of investment protection can factor heavily into extending the lifespan of equipment and helping to lower the TCO over several years of operation without sacrificing performance.
Integration costs can be significant when dealing with hardware and software components that are not designed for maximum interoperability. To mitigate the risk of unanticipated expenses while integrating corporate resources into a single framework, select a standard solution with a record of success in rigorous industry deployments. Choosing a solution that has been tested and deployed under a wide range of conditions and situations can help minimize implementation difficulties.
The time-to-market (TTM) costs correlate directly to the degree of difficulty and the time involved in bringing a new platform to full serviceability. Open architectures tend to reduce TTM costs due to the ready availability of components to support the infrastructure and the prevalence of skilled practitioners to configure and deploy the components. Proprietary architectures are generally characterized by limited options and reduced availability of specialists trained to configure and deploy systems, leading to a less-competitive set of market choices at greater expense.
Over the life cycle of an enterprise server platform, certain factors will have a prolonged impact on the intangible costs associated with that platform, and the combined effect of these operation costs can be considerable. The overall usability of a system affects everyone who makes use of the resources and accesses data residing on that system, but the direct financial costs aren't generally amenable to straight-line calculations.
Four other key factors speak to the overall utility of a system, and those factors that can influence operational costs in a number of different ways; they are sometimes referred to using the acronym RASM:
- Reliability – reliability of the system, measured in Mean Time Between Failures (MTBF)
- Availability – uptime of the system or application, measured in parts per million (PPM) of downtime
- Serviceability – easily restoring the system after a failure, measured in Mean Time To Repair (MTTR)
- Manageability – the ease with which the entire system can be managed, measured in systems per headcount.
Organizations where downtime can adversely cut into profits consider RASM to be an essential component of a viable application platform. In environments where a few minutes of downtime can cost an organization tens or hundreds of thousands of dollars, RASM can mean the difference between a thriving business operation and a business that struggles to maintain customer confidence and integrity.
By performing TCO comparisons, enterprises can factor real-world considerations and overall operating experiences into their assessments of the full range of expenses involved in launching a business product on a server platform. TCO costs include both tangible and intangible costs, and often the intangible costs can contribute significantly to the total expense of operating a platform.
Support costs represent a much higher percentage of the total expenses associated with a system than is often assumed. Under some circumstances, support costs over the life cycle of a server platform can exceed the original hardware cost by a factor of four or five times.
About the Author
Yasaman A. Ghazizadeh is an engineer with the Platform Enabling Group within Intel Software & Solution Services in Hillsboro, Oregon. Her current work is with the Core eTools team.
- PolyServe Total Cost of Ownership Calculator*
- Computational Fluid Dynamics Application using Fluent* Case Study of Fluent Inc.: Migration to Intel® Architecture Speeds up Modeling and Lowers Costs
- Intel Business/Enterprise Computing