Predictive Analytics Interview Series: Sarah Holder of Duke Energy

By: Eric Siegel, Founder, Predictive Analytics World

In anticipation of her upcoming conference presentation, Its Not a Black Box! Explaining Predictive Models to the Masses, at Predictive Analytics World San Francisco, March 29-April 2, 2015, we asked Sarah Holder, Senior Market Research Analyst at Duke Energy a few questions about her work in predictive analytics.

Q:  In your work with predictive analytics, what behavior do your models predict?

A:  Besides being a utility, Duke Energy also offers Warranty and Energy Efficiency Programs.  In the Marketing Analytics Group, we currently use Predictive Analytics to target Direct Marketing offers through Direct Mail, Email, and our Call Center.

Q:  How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A:  Predictive Models drive our Smart Window.  This Window pops up when a Call Center Representative pulls a Customers information.  A customer may be calling in to star service at an account, or to inquire about a change in their bill amount.  In the Smart Window, there is a list of top products for which the customer qualifies and star ratings to signify the probability of the customers interest.  Because the representatives have information on the customer, it helps them make a knowledgeable sales pitch.

Q:  Can you describe a successful result, such as the predictive lift of your model or the ROI of an analytics initiative?

A:  Duke Energy previously used external vendors for Targeting Direct Mail lists.  Once we moved the process to our internal Analytics group and used Logistic Regression to predict the best customers, our Load Control programs direct mail response rates increased by 261%!

Q:  What surprising discovery have you unearthed in your data?

A:  Contrary to a previous belief, we found that households with two occupants tend to use more energy as a whole when compared to households with 3 or more occupants.  The number of people in the household may be a clue towards the life stage of the occupants.  When there are children present, the home may not be occupied as frequently during the day, leading to lower energy use overall.

Q:  Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A:  I will briefly explain the success story of predictive modeling within our marketing department.  I will also review some tips for communicating the benefits of predictive models verses purchased segmentation systems for targeting customer direct marketing lists.

Dont miss Sarah Holders conference presentation, Its Not a Black Box! Explaining Predictive Models to the Masses, at Predictive Analytics World San Francisco, on March 31, 2015, from 3:55-4:15 pm. Click here to register for attendance.