For the next couple of weeks, I'm going to be writing a lot about derivative measures in social media. For me, a "derivative" is basically a measure that doesn't mean anything on its own - it's based on another, underlying measure, and is by itself meaningless.
A "derivative" is also an investment term, and in recent times one that has developed a bit of a nasty connotation. A derivative by itself has no value - its value is based upon two things: the value of the underlying security (like, say, a toxic mortgage), and the two parties involved in a derivatives transaction agreeing that the derivative is worth trading. When the world got into its toxic asset mess, it was because so much "value" was locked away not in the actual buying and selling of subprime mortgages themselves, but in all the derivatives that were based upon the value of those mortgages. In essence, they compounded the sins of the original investment exponentially, and when the mortgages became worth less, the derivatives based upon those values became worthless.
I think there are a lot of similar derivatives in the ways we measure social media. When companies commit to social media engagement, they naturally want to be able to measure how effective those efforts are. Because those efforts are throwing off a crapton of unstructured data to mine, and accessing that data is relatively cheap (if not free), there is seemingly no end to the number of "metrics" being created to measure these activities.
Recently, at the New Marketing Experience summit in Chicago, I was asked about the value of many of these measures, and I'll share with you what I shared with them: a given metric is meaningless until you can prove that it isn't - in other words, just because you can measure some kind of unstructured data doesn't mean it has any ties to the metrics that matter for your organization: Awareness, Trial, Churn, Hit Rate, Customer Satisfaction, Loyalty. Is a brand mention tied to awareness? You can't really answer that unless you have a sense of what your brand awareness in that channel was before and after the mention. A retweet is a derivative based upon a brand mention. "Influence" is a derivative based largely upon retweets. You see where I am going with this, surely.
When you measure your social media efforts in terms of mentions, retweets, "shares," "likes" and so on, you have to be honest with yourself - you are making assumptions that these measures have some value. They very well might, and they probably do, but you have to do the work necessary to prove it. At the NME Chicago summit, Discover's Steve Furman talked about tying Facebook behavior to new credit card signups. How did they do this? To quote Steve directly, "Brute Force." They didn't make the assumption that a Facebook fan had an intrinsic value - they did the work to tie a measured behavior to a metric that mattered - new card signups. Without that underlying measure, the derivative fan or follower measure is meaningless.
Similarly, your organization has to do the work before you ascribe any meaning to your social media metrics. This means measuring online - and quite possibly offline as well - before, during and after the activities you are trying to gauge. Is this difficult? Potentially, but it isn't a mysterious black box - it is a knowable unknown, to quote the poet Rumsfeld. If you do this work correctly once, you don't have to do it again for a good long while. If you don't do it, try reporting to a sharp CMO that your efforts generated 2,000 mentions and 1,200 retweets and see where this gets you. This will work for a while - especially while social media marketing activities are still part of a company's experimental marketing spend, but don't count on it lasting forever.
Please don't misinterpret any of this to mean that I think these measures are valueless. (OK, maybe "Influence" is valueless.) Do take this to mean that as the head of marketing for my own company, I expect myself to rise above lazy thinking, assumptions, and undisciplined effort - and I expect my direct reports, partners and vendors to do the same. You should, too.