I grew up observing and describing people’s facial expressions and body language to my father who was legally blind. The most fascinating thing to me was disconnects between people’s words and actions. This fascination lead me to study pattern recognition methodologies such as mixture and latent class models.

My work lies in the intersection of psychometrics and econometrics. It takes both perspectives to deeply understand and predict human behavior. I started thinking seriously about emotions and unstructured data while consulting to P&G in the mid-2000s. This was before Facebook, Twitter, and Instagram, so we were looking primarily at blogs and forums. I was involved in early “test & learn” projects that explored ways the “new” unstructured data from social media could be used to augment or replace brand and tracking research for the big P&G brands. Through this experience I started to see how social media could change market research. We were able to quantify the impact of sentiment, emotions, and topics people talk about online to brand equity and sales.

Over the past decade, I have been focusing on analytic and predictive methodologies which encompass emotional, attitudinal, and behavioral inputs extracted from unstructured data streams like social media, blogs, emails, and survey verbatims. Currently, I am experimenting with blending structured and unstructured analytical approaches. I guess you could say I was a data scientist before Harvard Business Review said it was fashionable!

Emotions drive human behaviors and a deep understanding of how emotions drive behavior can help you predict peoples’ behavior. Let’s talk about how we can work together so that you can get a deeper understanding of your customers and a higher return on your marketing investment.