Predict churn by mining behavioral patterns in telecommunications data
Every mobile operator has a wealth of information about its customers: their calling and spending patterns, location, handsets in use, demographics. But only a few operators manage to use this data effectively to make data-driven marketing decisions.
Churn Prediction by Exacaster
Exacaster is a specialist in churn prediction for Telecommunications companies. We have developed a unique approach that significantly simplifies the creation of very accurate predictive models with best in class performance. Create models with a few mouse clicks and that’s it – the heavy lifting of data preparation, model selection, model operation and performance evaluation is handled automatically by Exacaster.
Scalable Churn Scoring for Giants
Exacaster churn scoring process is implemented on top of Hadoop, the industry-leading Big Data platform. This ensures that we can scale our churn scoring process to very large customer bases without prohibitive cost. The largest Exacaster deployment to date predicts churn in a customer base of 10 000 000 customers.
Cloud Service for Startups and Niche Telcos
Exacaster operates a specialized Churn Scoring Cloud environment that allows even the smallest telco or MVNO to enjoy the benefits of churn scoring. The smallest Exacaster deployment predicts churn in a customer base of 40 000 customers.
The Business Case for Churn Prediction
Consider the following scenario: a mobile telecommunications operator has one million customers with 10 EUR ARPU and 2.4% monthly churn. How much additional revenue could you win if you managed to reduce your monthly subscriber churn by just 1/10?
The business case for reducing churn by 10% is thus very simple:
We consider churn reduction to be one of the key opportunities for every mobile operator or MVNO, especially as many markets have matured and protecting existing revenues is a key priority. A small reduction in churn rate gives a very significant impact on the customer base and revenues over a year’s time.