Fighting Churn With Data

Fighting Churn With Data

Fighting Churn with Data is the title of an upcoming book by Carl Gold, Chief Data Scientist at Zuora (NYSE: ZUO), about how to use data science techniques to reduce churn in subscription services. The book will be published in late 2019 by Manning Publications. In the meantime this site will preview the book chapters in my blog.

Latest Update: Code!

Fight Churn With Data is Now Open Source on Github! Check out the code…

Get Churn Right The First Time

Because there might not be a second chance!

Churn is the bane of subscription products and channels, and Data Scientists and analysts are commonly called to the rescue in helping companies understand what causes subscribers to churn, what is predictive of churn, and what can be done to reduce it.

While examples of churn are available in benchmark data sets, analyzing churn in the real world is fraught with challenges and pitfalls for the novice data scientist or analyst. Churn analysis begins with constructing a data set combining multiple raw data sources, and only ends when the findings of the analysis are communicated to field representatives in a way that allows them to actually reduce churn.

That’s right  – a predictive AI system alone doesn’t cut it when churn is at stake: To reduce churn you need to put your scientist’s hat on, test hypotheses and communicate the result to the non-technical crowd!    But don’t worry: The book and this site will cover every phase of the process, including the most dangerous pitfalls and case studies illustrating common findings.  The information here will allow anyone with only a modest data analysis background to get churn analysis right the first time.


Understand Churn

The first step is to understand the reasons for churn

Predict Churn

With understanding, churn can be predicted in advance

Reduce Churn

Churn reduction works, but you might only have one shot – don’t waste it!

Learn more…