MPP Global Launches Predictive Churn Functionality
MPP Global adds further enhancements to eSuite, the only fully integrated cloud platform to identify, engage and monetize digital audiences, with the introduction of predictive subscription churn functionality. This functionality has been added to the recently launched Retention & Recovery module.
The technology helps organizations identify high risk subscribers who are likely to churn enabling them to proactively target and re-engage with these customers in order to minimize churn and maximize revenues.
Predicting Subscription Churn
The predictive subscription churn functionality employs Machine Learning techniques to apply a churn risk probability to each active subscriber. The process analyzes customer behavior to generate reports of customers who are most at risk of voluntarily cancelling their subscription. The performance of the prediction engine is transparent with the previous month’s churn prediction accuracy available for analysis. As clients use more of eSuite’s modules and functionality, the opportunities with the scope and precision of the prediction engine increase.
Machine Learning & Predictive Engine
Machine Learning is a technique where a highly scalable processing engine, historical consumption, financial and consumer data, along with specialist algorithms are used to estimate the churn risk for every active subscription. Using a minimum of 6 months of historical data, the business defines set data dimensions containing information about previous churned subscriptions which may indicate increased churn risk. The advanced algorithms are then ‘trained’ using this data.
The high processing capability of the machine learning cloud platform is used to process current active subscription data, outputting churn risk probability for each active subscription. This process is typically repeated each month resulting in new target datasets of high churn risk subscribers delivered on a regular and automated basis. In live environments where predicted churn was measured against actual churn, precision rates of 85% to 98% were experienced. Precision rates are a measure of how many subscribers, which are identified as high risk, actually go on to churn in the following month.
Using historical data to predict when a paying customer is likely to voluntarily churn enables businesses to react and automatically target them with an offer or incentive in order to retain their custom. By targeting them with a promotion or engagement activity, companies limit any damage to the customer relationship and can increase customer retention. It is also possible to build business cases around campaigns, based on precision and conversion rates, ensuring retention campaigns make commercial sense.
Advancements and Profiling Data
Machine learning technology is playing an important role in developing eSuite’s capability. It is being employed in scenarios such as analysing the data of existing customers to understand at which stage of the journey they purchased and which product or service they are most likely to buy in the future, based on previous behavior and profiling. For example, if an anonymous user clicks 5 sports articles or videos in succession from a mobile device, eSuite can recommend a product or service which best matches their profile.
This advanced technology provides businesses with insight into their services which they have not had previously. It enables the opportunity to dynamically adjust the user experience based on their behavior and preferences. Businesses can then act on the front foot and reach audiences who may not have known about a particular product or service, driving subscription revenues and further developing customer relationships.
MPP Global Commentary
Playing a lead role in the project, Chris Cheney, CTO at MPP Global, said: “We are very excited with eSuite’s predictive churn functionality and the benefits it will bring to the market. We have the expectation to bring in further data points for which clients will understand more about what leads to subscribers churning and act in the best way to prevent it.
“Churn prediction is just one use case where machine learning is used, with many more planned around the use of advanced analytical tools and capability. You simply can’t do this stuff unless you have the platform components to be able to collect and process the important data and our product strategy has very much been around the positioning to enable this.”
Paul Johnson, CEO at MPP Global, commented: “This project has been a major development within MPP Global and redefines what we are able to offer clients. eSuite is already the only fully integrated cloud platform to identify, engage and monetize audiences and now with predictive churn we are truly offering a one of a kind solution which will help any organisation offering subscriptions.”