A better way to measure true costs

Last week we discussed the concept of total cost of service delivery. In a nutshell, it is an approach for measuring the value of technology based on the assumption that you’re putting the technology in place to help accomplish a business function.

Why add a new buzzword when we’ve already got tried-and-true total cost of ownership (TCO) and return on investment (ROI)? Because, in my experience, neither model adequately captures the true effect of technology on a business. TCO is great if you’re worrying about running your IT department less expensively: A product that requires two engineers to manage is clearly an improvement over a product that requires 10. Unfortunately, this approach doesn’t tell you whether buying that product was a good idea in the first place. ROI tries to address this weakness by asking about the return on a given technology investment, but there are precious few clues as to how to quantify that return.

Enter TCSD. The fundamental concept is that technology – even infrastructure technology – is deployed in the service of a particular business function (or “service”). To measure the value of the technology, look at the role it plays in performing that function.

An example: An accounting service might require two accountants, a specialized analytics package, hardware and software, and a network. But you might be able to offer that same service with no accountant, no specialized analytics, and just the hardware, software and network – clearly an improvement in TCSD.

Some guidelines for computing TCSD:

1. Start by defining the service using business terms. It’s not enough to know that FTP makes up 12% of your WAN traffic; you need to ask what applications are generating that FTP traffic and why. Is the research and development department sending CAD/CAM files? A switch updating its call records? Find out who in your business owns that application, and open a dialogue.

2. Understand the performance parameters of that business function. Do files have to be updated hourly? Why? What happens if they aren’t updated on that schedule? Would the ability to perform synchronous updates be an improvement? Take your time with this step, and be creative. Many times, business units do things a certain way because they’re unaware that better technology can provide improvements.

3. Translate the business performance parameters into technical parameters. If the goal is to perform synchronous updates, what does that mean in terms of megabits of bandwidth and milliseconds of latency?

4. Compute the current cost of providing the current service by including infrastructure hardware and software, business applications, and personnel. Be sure to include costs that are borne by the business unit (not just IT): the cost of a business specialist, for example, or specialized software.

5. Now you can start your modeling. Is there a different combination of hardware, software, services and humans that could deliver the same service at a lower cost or measurably improve the service’s quality? Offering synchronous updates at no extra cost might directly benefit an organization’s top or bottom lines.