Drexel prof’s new book: for marketers eyeing the power of data science

Fifteen years ago, when garnering data about customers wasn’t as sophisticated as it is today, Elea McDonnell Feit jumped right in to the emerging field.
The Kennett Square native — now an assistant professor at Drexel University and on Wharton’s Customer Analytics Initiative team — said she’s always had an interest in understanding customers and solving problems.
“When research started, all we had was surveys. People would fill them out and mail them back. That’s how we got data about customers,” Feit said. “But research now has focused on advertising response and using math to understand how to solve problems. It’s completely undiscovered.”
Feit currently works on projects with companies who send out emails to gin up business. The question at hand, she says: are they more likely to buy just because they get an email?
Last year, Feit teamed up with a colleague who also enjoys using math and analytics to understand people. Together, she and Chris Chapman — who now works at Google — wrote R for Marketing Research and Analytics, a book for marketing professionals coming out this spring.
Pre-order for $52.30
R, the statistical programming language, is widely used among statisticians and data miners, Feit said. Its open source nature allows users to download, share and add their own code, which has become a popular outlet for top academic statisticians to contribute their work to a larger concept.
But marketers (also a group that relies heavily on data) have been slow to adopt the R language. That’s why she and Chapman wrote the book.
It’s aimed at marketers who want to do more with data science, she said.
“Chris’ goal for the book is to get marketers over the hump and to hook up to big R data systems,” Feit said. “The book is designed to type every line of code and to push you to learn code.”
Beyond the new book, Feit said she’s enjoying focusing her work on something she feels is important: data-driven marketing.
“All of the sudden, if you’re a hairdresser with 200 clients — once you get to any larger scale, you need data science to get you through that,” Feit said. “It’s about thinking strategically. If you send out a catalog, what past behavior should you look at to inform who you send it to? It’s about pulling out of this data mine.”