The Battle Over Metrics: Is ROI an appropriate measure for social media and brand?

My new position in the not-for-profit world has had me thinking a lot about marketing measurement and not only how to prove "success", but also how to best allocate funds across different initiatives in order to get the maximum gain (in my case, for the organization). How do you determine that if you pulled some funding from one initiative in order to allocate it to another that you would end up with a stronger ROI? And what about the areas of marketing that are not as easily "measurable" in terms of ROI, like social media, for example? How do you decide to pull money out of the marketing budget fot these areas over "tried and true" direct mail campaigns, etc.?

Through the wonders of the internet, I found an article that discusses just this topic. It was written by someone by the name of Ron Shevlin , in his blog Ron Shevlin's Marketing Whims.

Shevlin refers to social media, as well as branding, as "infrastructure", and so when we allocate funds for these endeavours, we are in fact investing in infrastructure. This is an interesting approach. I suppose the same could be said of PR. Shevlin states that these areas are often measured in terms of KPIs (key performance indicators), but that these are not adequate performance measures. Shevlin claims that these are not good indicators of incremental changes in performance or of their effect on the whole marketing funnel. I agree. I think they are also rather arbitrary.

This puts online marketers, branding folks and the like on the back foot when they try to explain their efforts to marketers obsessed with ROI (like me). Afterall, what is the ROI of having a company Twitter account? You tell me. You see, I think it all boils down to the fact that consumers (be they companies, donors or individuals) are complex creatures. They (we) don't make buying decisions based on only one factor. So while logically most of us would say that, "yes", the brand name was important when we bought our MP3 device, we would be hard pressed to report how influential that brand was in our buying decision or whether if presented with another item of the same value, but weaker brand we would still make the same decision.

At the same time, I don't buy it that social media, PR, WOM and all the other forms of "alternative" marketing (I can hear my colleagues in these fields screaming at me now...), can't be measured. Sure, maybe these areas are harder to measure than say, direct mail, but surely not impossible. Okay, so what if we can't have perfect forms of measurement? We should still be able to show progress. We should collect what info we can. We should watch for trends. And with enough data, surely, surely, we should be able to show some correlation between these trends and ROI. Then we can take a crack at determining how much of our marketing budgets should go towards "infrastructure." Don't agree? Comments are open, folks.