Listen to Your Metrics – They’re Trying to Tell You Something

July 23, 2008 | Posted by: Thea Roberti

Within the last couple of weeks, there has been much written on causation versus correlation analysis and the weaknesses associated with both. See Salon's Lies, damn lies, correlation and causation and KD Paine's blog post in response to this piece.  After years of fighting what seemed like a lone battle against using causality in an attempt to pump up PR ROI, I am so relieved to finally see this topic getting some attention.

As the commentary underscores, we can't say that media (social or traditional) causes revenue, because there are so many different reasons why revenue occurs. And in my opinion, even correlation can by sketchy at times. As a researcher (a scientist at heart), I don't like sketchy. Sketchy has no place in the world of measurement - at least not in my measurement. Why do we need to resort to smoke and mirrors when the basic metrics alone provide a wealth of information?

Mike Moran does a phenomenal job of describing this in his post Biznology: Why Marketing Needs to be Scientific. I love Mike's argument for letting your metrics be your guide: "When you do something really cool and the metrics tank, stop. When that campaign you just love fails, kill it. When the dumb idea the intern came up with takes off, do more of it. If you regularly feel embarrassed and stupid about what the numbers tell you, then you're doing it right." Mike concludes that you should save your brain for analyzing what the metrics say, rather than doing correlation analysis. I couldn't agree more.