Member Renewal Model for Optimizing the Marketing Strategy

Case Type: Develop a renewal model to categorize members on their likelihood of renewing their membership.

Partner Type: One of the largest Sports Event Live Streaming companies.

Challenge: To identify the cohorts of members that are likely to renew their membership and tailor the marketing strategies to optimize the marketing expenses.

Solution: iSOCRATES Managed Services’ Machine Learning model to categorize the members into low, medium and high risk of not renewing their memberships.

Approach: The iSOCRATES team leveraged the available data to address the business problem.

Initially a comprehensive EDA (Exploratory Data Analysis) was performed to validate the data to check if they were consistent with the business rules. Post analysis, iSOCRATES corrected all identified database anomalies.

Upon EDA, the team was able to identify the key drivers of membership renewal. These key factors were used to train a robust machine learning model.

The model was evaluated using the metrics like precision and recall. An iterative process was carried out for the hyperparameter tuning to arrive at a model that generalizes the problem well without any bias.

Our model’s precision was 83% and recall was 94%. This is tracked monthly during the prediction process every month on a new set of incremental data.

Apart from the member categories, we also provided the key insights and actions using MADTECH.AI, that can positively impact the renewal rate.

• Key drivers of renewal rate
• Members categories like high, medium, and low risk of not renewing
• Engagement behavior of the members based on website visits
• Potential member targets to increase website visits and subscription

Results: Instead of having a uniform marketing strategy across the member universe, iSOCRATES’ Renewal Model provides the cohorts of members based on the renewal probabilities, which can be leveraged to have different marketing strategies for different cohorts of members, thereby reducing/optimizing the marketing cost and improve the renewal rate of the members and memberships.

On average over the past 6 months, 93.7% of the members who have been classified as medium and low risk members have renewed their membership within a year from their previous membership expiry date. This shows the efficiency of iSOCRATES’ Renewal Model.