Physicist cracking the marketing funnel for meal-kit delivery company - Propheto Case Study

Published: January 8, 2021


Ben spent his PhD literally staring at the stars. His work in astrophysics involved telescopic data, studying the light emitted from dying stars millions of lightyears away.

But now, nearing the end of his PhD, Ben realized he wanted to work in industry. He was ready for more applied work where he could use his skills to generate immediate impact:

"As a science PhD student planning to transition into data science upon graduation, I was very excited to discover that I might be able to gain valuable industry experience while still in school."

The challenge was that he had no practical experience in industry. Throughout his PhD program, Ben never had the opportunity to even explore let alone experience working for a company.

When Ben found Propheto, he was excited that he could finally apply his skills on a real project and earn his stripes that could help him land a full-time job as a data scientist when he graduated. Also, making twice his salary as a TA was an added bonus (pun intended)!

Soon after signing up, a project caught his eye. A home meal-kit delivery company was looking for a data scientist to help them understand how their customers converted to subscribers. Would customers mainly choose to sign up because of a Google Ad or a blog post? How many ads did a customer see before they signed up? Answering these types of questions required someone with strong statistics and programming skills. Ben was interested because the problem was unique and would make a material impact on the business:

"This sort of project is of interest to me because it involves sifting through large data sets and finding clever ways to link them together, and of course, the results of my analysis have the potential of directly increasing revenue while reducing costs.”

While on the surface, this project seemed way out of Ben’s experience, it actually aligned with a lot of the work Ben had done. Ben was accustomed to collecting and cleaning messy telescopic datasets from different systems for his analyses and applying statistical methodologies to uncover trends. That experience translated well to this project, which required merging and cleaning messy data from different advertising systems to determine any impact on a visitor's likelihood to subscribe.

After speaking with the company’s Head of Analytics, Ben was matched on the project and began working remotely. Ben was able to help the Head of Analytics by offering more sophisticated data munging and statistical analysis techniques, skills the team lacked and desperately needed for this project. Within two weeks he was already generating impact:

"In the process of extracting relevant data for my project, I have uncovered a few places where the existing database infrastructure is not operating correctly or as expected. As a result, the data team has been made aware and the problems will be fixed. For me, this has been a nice way to add value that is distinct from the main objectives of my project."

Through Propheto, Ben gained valuable experience applying his data science skills to an actual company problem. By getting outside the academic context, he demonstrated that he could effectively apply his technical knowledge in a completely different environment and successfully be an industry data scientist:

"Propheto provides PhD students like myself with access to interesting data problems that actually matter to the companies that pose them. While mentorship is there when I need it, I truly am a consultant rather than an intern, and the distinction means that I have a similar level of autonomy in how and when I do my work as compared to my responsibilities as a graduate student.”