On Storks and Babies: Correlation, Causality and Field Experiments

Abstract

The explosion of available data has created much excitement among marketing practitioners about their ability to better understand the impact of marketing investments. Big data allows for detecting patterns and often it seems plausible to interpret them as being causal. While it is quite obvious that storks do not bring babies, marketing relationships are usually less clear. If marketers want to be sure they are not walking into a causality trap, they need to conduct field experiments to detect true causal relationships. In the present digital environment, experiments are easier than ever to undertake, but they need to be prepared and interpreted with great care in order to deliver meaningful and genuinely causal results that help improve marketing decisions.