Customer Churn

The Challenge

The challenge for many Sales and Marketing organisations is predicting when customers might cease buying your product or consuming your service.  Whilst much time and money is often spent by organisations acquiring customers, there is a growing realisation that retaining customers requires significant focus and investment.  

Universities, Retail, Banks, Telco and Utility organisations, to name a few industries, have all invested heavily in leveraging data platforms to help identify patterns that might indicate that a customer is about to lapse or move to another provider.  

With Data Science becoming more mainstream, and tools and technologies becoming readily accessible, most organisations are now looking at understanding how Customer Churn impacts their business operations.  

Our Learnings

Our experience with implementing predictive models to uncover patterns in Customer Churn has highlighted a number of learnings:  

  • Defining how customer churn impacts your organisation is often the first important step in identifying how to solve this business problem;

  • Identifying the right sources of data and the features and attributes required in your model can be a complex process requiring many stakeholders to be involved (eg. Sales, Marketing and Finance);

  • Its important to start small and incrementally develop a predictive model by training, testing and validating the model with the right volume of data; 

  • Identifying longer term trends in your customer data can help to explain data anomalies and outliers;

  • Expect to have contradictions and inconclusive results, particularly in the early stages of building the model;

  • Use a data analysis tool that enables a wide variety of data sources to be quickly ingested, modelled and reported within the tool.  To that end, we find Power BI is a fantastic tool to begin with.

  • We like to use 3 classifications to quickly identify churn observations - Churn Less, Churn More and Little/No Churn.  This then enables discussions to be held with business stakeholders around conducting further experiments.

Benefits

There are many benefits to be gained from building predictive models around customer churn:

  • Building a shared and strategic alignment across the organisation as to why customers churn; 

  • Segmenting your customers based on churn analysis will enable you to determine how you will sell and service to them;

  • Enabling your marketing team to target their marketing activities, particularly around promotions and offers;

How we can help!

Contact Revenite to organise a session to discuss how your organisation might implement a customer profitability framework using the latest data science tools, techniques and technologies.