I have long maintained that working in an IT department can result in a net gain of zero. Too many IT professionals spend most of their time doing two things: trying to stay off the obituary page and occasionally trying to make the headlines. An understanding of total cost of ownership (TCO) is a good way to make that front-page story and get noticed in a positive way.
TCO is receiving increased attention today as organizations realize that distributed computing operations are expensive and require a strong, integrated set of management processes. A TCO study can provide the necessary structure to begin building those processes. A proper study can also serve to make senior management aware of the complexities of distributed computing and help justify budget and staffing requirements.
Developing a specific TCO for your business is one way to make the budgetary case. Your finance department should be included in the process to define the cost of ownership so it can see the reality of distributed computing operations and increase the likelihood that it will become a strong ally for supporting the additional
IT funding needed to better manage expenses. In short, without a clear understanding of TCO, an IT organization faces eventual disaster and possible extinction.
Having worked on TCO initiatives as a Gartner analyst, I often saw clients make fundamental mistakes in their analyses and come to erroneous conclusions. The key to a successful TCO study is to avoid the pitfalls. The way to do that is to understand the following common myths associated with TCO:
• You need to match someone else's numbers. No. Mileage varies, and enterprises should use TCO models to perform their own analyses. Don't take someone else's numbers as being valid for your organization. There are no standard TCO numbers that matter except the specific costs to your organization. TCO tools are a good starting point but only that. No one can tell you your costs except you. Remember that models are valid for creating assumptions, not final costs.
• First and foremost, TCO is about technology. Wrong. TCO is more about philosophy and best practices. Technology is an enabler, not an endpoint.
For example, Microsoft advertises that Windows 2000, as a platform, has a lower TCO than prior versions. Clearly, deploying Win 2k doesn't cause your laser printer to start spitting out $20 bills, so what's the secret? Easy. Win 2k simply has more out-of-the-box technology that helps with manageability, and it's those tools that help lower costs. But deploying a new operating system that promises a lower TCO, without enabling the management features that actually control costs, can actually lead to cost increases.
• The platform with the lowest TCO is always the best choice. Never. As an analyst, I was often asked what platform had the lowest TCO. My answer was always the same, "A yellow pad and a pencil."
The cost of a given platform might very well be expensive, but it needs to be balanced against the potential return on investment. Yellow pads and pencils are cheap but not when you need to collaborate with someone on the other side of the world. The platform with the lowest TCO may not be the best match for end-user requirements.
Lowering TCO is really about management philosophy combined with the technology that enables it. Lowering TCO doesn't involve shortcuts or magic.
If you're thinking about doing a major operating system installation or migration, start by evaluating tools and hardware that reduce costs, such as remote or unattended installation and setup. Also, consider the use of policy managers for greater consistency and supportability, as well as software healing-and-repair tools for problem resolution and recovery. Avoid the myths and lower the costs.
Now go out and make some headlines.
Gartenberg, former vice president and research area director at Stamford, Conn.-based Gartner Group Inc., is looking for the next generation of Internet technologies. Contact him at michael.gartenberg@mindspring.com.