a couple weeks after starting at tresata, i went to grab some coffee with one of my co-workers, Jack. while we were waiting in line at Starbucks, he asked me “in what ways could Starbucks benefit from knowing how much a customer paid as a function of how long they stayed there?” my first impression was to create a duality between the coffee-to-go people, and the loungers. loungers can then be broken into people socializing, reading, working on a computer, etc. people who are socializing could be there for business, friends, people watching, etc. if an array of specific consumer data were present, Starbucks could have an opportunity to gain unprecedented insights into their consumer interactions.

i cannot claim to have solutions to creating new avenues of profit for Starbucks. i’m also not claiming to be experienced in customer intelligence. i come from a physics and computer science background, where my rational constructs tend to fall short of being useful in an end-to-end analysis. however, the skills developed through the position of a data scientist are applied in response to the complexity of identifying opportunities with understanding consumer behavior. therefore, it necessitates the creation of a new bridge between algorithm development on distributed systems, and domain expertise that can help extract insights from the vast mounds of data. it becomes the exploration of a process to ask larger questions and to see past the stereotypes of ‘coffee-to-go people’ and ‘loungers’. for the curious, it is a profession to see a solution propogate from esoteric statistics all the way down to a reified solution…even while waiting in line for coffee.