Katie Paine has conceded a point to Shel Israel: As cheap as they are to set up and maintain, there’s really no need to measure the ROI of blogs. After all, is it necessary to measure the ROI of your pants? (Shel denies he ever used this analog, but Katie remembers it.)

There’s much to be said of this argument. If a company installs the open-source Wordpress on its own servers, the cost is limited to bandwidth. For $14 a month, a company can have the top-tier Typepad service without concern for bandwidth use. When measured against cost, it doesn’t take much to justify. One minor positive outcome can tip the scales and rationalize the cost of the blog.

It’s even easy to extend the “pants” argument to higher costs, as Toby Ward does when discussing intranet ROI. Intranets, obviously, cost more than blogs, but Toby has asked if companies demand ROI justification for their telephone networks? Not even the bean counters insist on a tally of the ROI for phones because everyone knows the consequences of removing them.

But I would make the argument that the kind of ROI assurance executives want is not the kind of ROI accountants measure; we’re getting too hung up on definitions. The ROI in question here is more commonly referred to as a cost (or risk) benefit analysis. While ROI generally refers to the amount of money earned after all expenses are tallied, the risk-benefit analysis weighs the consequences of an action or expense against the benefits it will achieve. Because so many executives fear the consequences and don’t understand or recognize the benefits, I still think this kind of measurement is important.

What’s not important is a dollars-and-cents accounting. At the New Communications Forum, Charlene Li raised the story of the UK blogger who lamented the inability to find his favorite deodorant and the way Unilever was able to capitalize on the post to boost its reputation. That’s an example of the benefit of monitoring the blogosphere that outweighs the risk (the time required to conduct the monitoring and its subsequent drain on productivity).

The employee-written blog from Intuit’s Quickbooks Online that resolves customer concerns equally justifies the amount of time employees spend generating content for the blog and offsets the concerns that an employee might say something wrong. After all, what’s the value of turning a disgruntled customer into a satisfied one who sill sing a company’s praises to friends, family, and (possibly) the audience of his blog readers?

Not all measurement is ROI. Anecdotal evidence counts as measurement as much as numbers do; it’s why measurement experts include both quantitative and qualitative measurement tools in their toolkits.

But until management sees blogging as indispensable as the phones—or their pants—I’ll continue to recommend that the benefits and outcomes are measured, even if only anecdotally. Katie concedes that there are some things that don’t need to be measured. I agree: Pants and telephones are two great examples. But executives aren’t arguing that pants are unnecessary or that telephones pose unnecessary risks. For skeptical executives, measurement remains the most convincing and compelling tool for changing minds.