As individuals who are passionate about the solar industry, gathering meaningful data, and creating holistic strategies, we were immediately drawn to the Tesla #1 challenge of identifying appropriate potential solar clients.

Investigating and testing various aspects of the data provided gave us directionally correct theories of likelihood of "going solar", but we felt there was something missing. After all, from a solar installation/financing perspective, lead generation is the bread and butter of business, and lead conversion isn't just indicative of the bottom line, it is the bottom line. And at the end of the day, a lead, no matter how promising or accurately modeled, is only a success after installation. And it's not homes that cancel leads, it's the home owners.

So we asked ourselves, how can we get more insightful data about home owners? Is there even a way to understand and qualify personality traits of humans and their propensity for going solar? We thought.. yes! This unearthed a personality quiz with unobtrusive data insights derived from fun and immediately engaging questions. These personality questions helped to identify traits such as consumer patience (after all, the installation process can take upwards of 120 days), as well buyer's remorse (a 30k purchase is nothing to sneeze at).

How did we do it: Initial data review using tableau, pivot tables Joined datasets and cleaned information Sklearn random forest classifier K fold tested training data Eliminated outlier features Discovered important attributes: Attributes like X,Y, Z played a focal part in the predicted likelyhood of completing the process

Share this project: