Sprocket used quantitative research and machine learning to turn STANCE personas into actionable customer segments.
When other apparel brands zig, Stance zags. The founders at Stance saw apparel categories (socks and underwear) that had been ignored, taken for granted, looked over, and dismissed. They decided to infuse fresh design in their products and ignited a movement of art and self-expression as a result.
Stance came to Sprocket looking for help in increasing customer lifetime value (LTV) by encouraging repeat behavior. A fundamental question was tackled as a part of this business challenge: what did Stance customers look like when you take a close look?
To answer this question, the team at Sprocket combined quantitative research with machine learning to discover unique Stance customer “personas”. The first step involved co-creating an online survey with Stance that would get at the motivating factors and purchase preferences of customers. Once that survey was complete, unsupervised machine learning (a fancy phrase for letting machines tell us what types of customers we have) combined with smart Sprocket hypotheses were used to identify unique groups of customers.
The combination of a behaviorally-focused quantitative survey and machine learning created the unique advantage of being able to understand customer motivations and needs (to better market to them) as well as the ability to identify each customer in the Stance database with a probable persona. This turned a traditional marketing persona into an actionable marketing segment that could be used to target offers and messages to increase customer lifetime value.