This article outlines the datasets required to use the subscription option in the AI & Analytics Engine's customer churn template.
The Customer Churn Prediction template provides a complete machine-learning solution for predicting the likelihood of customers churning in the future.
To create an app using this template, you’ll need to provide enough data about the customers' activity during a relevant period in the past.
💡The required minimum period of time for the recorded customers activity is determined by the AI & Analytics Engine based on the given app configuration. It displayed in the Add data stage of the App building pipeline.
🎓 For more information about the different time spans showed, read how does the Engine process data for the subscription option in the customer churn template?
Add data stage of the subscription churn template
In the example above, the minimum period of recorded customer activities for the template to work is 231 days.
The models can then learn to predict churn by analyzing the customers' data, by looking at behaviors and trends that can distinguish churners from non-churners. For the subscription option, the template requires:
A dataset containing the subscription start dates and churn dates of customers who were active during the chosen period
At least one event log dataset, with each one containing a particular type of interactions between customers and the products/services offered by the business: Eg. voice messages, browsing sessions, service calls, etc.
Additionally, a customer information dataset can optionally be included in the template, containing
- Biographic/Demographic profile information: Eg. age, place of birth, current employment status, education level, residential address, household size, annual income, etc.