How Carl Gold Fights Churn with Data | Speedify LIVE

On the 170th episode of Speedify LIVE are joined by Carl Gold PhD, director of Machine Learning and Implementation at OfferFit and author of the book that’s been passed around the Speedify office, Fighting Churn with Data!

We chat with Dr. Gold about running unconventional streams on Twitch, the importance of Privacy, and how one should start if they want to get a job in data science.

Here are our 5 takeaways from our chat with Dr. Carl Gold:

1. Dr. Gold runs somewhat of an untraditional livestream on his Twitch channel: he does programming live! He says he’s tried both structured and unstructured streams, the latter doing worse, so he usually goes in with a plan. He generally starts the streams explaining some of the basic terminology like churn, then works on an example from his book. 
2. When it comes to Privacy and Churn, we learn that most private data is usually not that helpful to churn and can be completely anonymized! While some data, like demographic data is useful to things like predictive analytics, Dr. Gold says that PII is less important and can usually be emitted, and if it can be anonymized it should be! 
3. Dr. Gold says he originally wanted to write an advanced book in data science, but his publisher wanted the book to be more accessible to more people! Now, thanks to Manning publications, Fighting Churn with Data is something that could be used as your first data science course – although you do need an understand of programming in Python and SQL. 
4. When it comes to starting out with data science, there’s a few things Dr. Gold recommends first: have some understanding of basic statistics including econometrics, know the basic Python packages like SKlearn, and then choose a specific area of data science you want to focus on and highlight that on resumes before applying for a job.
5. Dr. Carl Gold’s stream setup is fairly simple: he runs the stream through Streamlabs OBS on a Mac, and he uses a Logitech camera for video and a HyperX SoloCast microphone for audio. One thing he highlights is that if he could go back, he would invest in a good mic sooner!