Proponents of Big Data as a tool for organizing information in business and in politics claim that, by using rich data sources such as social media, they are able to predict the preferences of customers and voters, targeting communications in an intelligent manner that simultaneously reduces cost and increases effectiveness. In theory, the ideas are exciting.
In practice, Big Data systems often end up looking like big idiots. This was the case for me this afternoon when I visited the Ready4Bernie Twitter profile. There, in the big bold form that is the signature of the new style of Twitter profile, I saw that Twitter’s Big Data system for predicting which Twitter accounts I should want to follow, suggested that I try…
… the Bank of America.
Why? What’s the connection between the likely presidential campaign of Senator Bernie Sanders and Bank of America?
Bernard Sanders is a frequent critic of Bank of America. So, the senator’s own Twitter feed, as well as Twitter feeds of his supporters, often include the phrase “Bank of America”. Twitter’s predictive system is too stupid, however, to notice that Sanders and his fans are describing Bank of America as heartless, corrupt, abusive, and vile.
Mature marketing systems recognize that context matters. Simply recording that people use a phrase isn’t sufficient.
Big Data, you’ve got a long way to go. You’re not even medium large yet.