Brand health tracking methodologies are facing significant headwinds. With increasing media fragmentation and pervasive social media use, researchers are really challenged to establish cause-effects using only big-tent, catch-all measurement approaches. We are fast-forwarding to an era when marketers will be pressed to use multiple data streams in order to triangulate on their real brand and communication issues . One such data stream gaining traction is the real time listening by mining the social web – preferably custom configured for the category.

Notwithstanding the fact that social media does not mirror all consumer opinions, the real time access to metrics such as share of voice on social media, volume and valence of buzz, from which source and the raw verbatims feels so powerful. They might be unrepresentative samples numerically, not influence-wise. Tropicana had to revert to its old design bowing to an online outcry fearing it’s percolating to a broad swath.

I have been debating this with the Global Analyst of Millward Brown. Traditional data capture methods such as online surveys are increasingly becoming unrepresentative. There are issues about reliability and validity associated with panel samples. Panel member engagement is waning, panels are thin on hard to reach groups (millennials, high networth, time-poor, etc.).

We need more rigor in constructs and measurements than before. Welcome to the world of complexity!