A manufacturing company fuels growth with an effective customer retention strategy by pro-actively predicting customer churn

A leading manufacturing company was struggling to analyze and accurately predict customer churn. They partnered with Applexus to transform their analytics capabilities and drive accurate churn prediction. With the increasing cost of acquisition (CAC), churn prediction is a critical business use case for enterprises looking to develop a strong customer retention strategy. Furthermore, the fact that high churn rates are far more likely to accrue over time posed a threat to the company's long-term potential growth.

Challenges

Our client needed to identify the predictors causing loss of business due to customer churn. This information was critical for them to redesign their marketing and promotional efforts.

• Although a system was in place to assess customer churn, mostly the analysis was conducted after the fact, consumed with hindsight bias. These informative findings were not enough to drive actionable insights to prevent customer churn.
• The client lacked a modernized analytics landscape to support advanced analytics projects. Enterprise reporting and business intelligence solutions were required to be transformed into AI-empowered advanced analytics solutions.
• A governance framework for implementing data science initiatives was absent. Added to that, there were data quality issues impeding the effectiveness of any analytics initiative.

Solution

Applexus engaged with the client to identify the predictors / influencers of lost business due to customer churn. Additionally, a user friendly platform was required to provide intuitive insights to the business users.

• Data lake and predictive analytics customer churn model were designed to deliver predictive insights to proactively address customer churn and gain a competitive advantage in the marketplace.
• Predictive visualizations and dashboards were prepared for each location or group to identify the key influencers of customer churn, including product categories/product category mixes that contribute to predicted attrition.
• Foundational capabilities for the design and deployment of advanced analytics use cases were developed. Data lake architecture was designed and aligned with the Data warehouse. This activity was supported by establishing proper governance, processes, delivery model, and methodology.
• Change management was also addressed through organizational impact assessment and detailed training requirements.

Results

Predictive modeling provided insights into future attrition rates for current customers and also the segment of customers that are likely to churn (by industry, size, and tenure). Moreover, the model also provided a time frame in which customers and/or products start to exhibit attrition signals. The client could actively visualize dashboards that displayed actionable insights to proactively attend to customers and product segments to reduce churn.