For example, swapping out one accounting system for another is a TCO project that while possibly important "is not going to win out over one that provides tools to my sales force," Harris says. Shaklee markets personal and household care products such as vitamins and cleansers. The prevalence of the ROI metric reflects the position of IT, at some companies, as a business department like any other, says Jeff O'Hare, senior vice president of IT at West Corp.

IT decisions should be less about technology and more about the business capabilities the technology makes possible, he says. Therefore, IT proposals should be measured like other business ideas, he says, for their potential to bring speed, efficiency and innovation. ROI offers a strong way to represent those ideas.

Sadin, at Loomis, explains that his projects are prioritized against proposals to, for example, open new branch offices or buy new trucks. "In every other part of the business, ROI potential drives every decision. IT can't have its own governance model."

Getting business units to fund IT infrastructure projects, which are frequently justified with TCO, can free up other money for technology innovation, CIO columnist Dean Meyer recently argued.

But remember: Neither ROI or TCO accounts for every cost or every financial benefit that may ripple from a project, says Erik Dorr, senior IT research director at The Hackett Group consulting firm.

Those metrics help companies choose how to spend their money. But in the decision ultimately made, "a lot of healthy business judgment is involved," Dorr says. "You can prove anything you want with a spreadsheet but good managers have good intuition in judging how solid their assumptions are."