BiddRocket is a behavioral science company that has been studying how groups of people make decisions since 2010. Along the way we discovered that hidden in transactional data are networks of social influence similar to the network data owned by companies like Facebook, Google, and Twitter. The big idea is that merchants do not need these large social networks to learn who their most valuable customers are. We have developed Contigo.io; a machine-learning cluster that infers social influence networks from transactional data. It reveals connections between customers and influence among them.
Basically, we are able to infer a social network structure only from transactional data logs e.g. sales data. More data is good, but all we really need for our technology to work is payment data. That means that we don’t need graph data from Facebook to tell you how your customers are related and who is most influential. With our definition of influence being that the most influential customers are the ones who have been proven to drive most sales among the others. Furthermore, don’t forget that even when Facebook graph data is available, online influence is only a shadow of real world influence.
The Contigo.io Dashboard allows users to upload large data sets for processing. After an influence network has been inferred, users can download the database in full or explore specific clusters online.