Subscriber growth forecasting in the streaming industry: A Netflix case study

Accurately predicting subscriber growth is critical for streaming platforms seeking to optimize customer acquisition and retention strategies in an increasingly competitive market. This study explores data-driven approaches to forecasting subscriber growth by analyzing a Netflix userbase dataset. Using a comprehensive methodology that includes data preprocessing, feature engineering, model training, and evaluation, the research investigates the effectiveness of three machine learning algorithms: Random Forest, XGBoost, and Gradient Boosting. Among these models, XGBoost achieved the highest performance, with an R-squared value of 0.987 and the lowest RMSE of 42.562, indicating its superior ability to capture complex subscriber growth patterns. To ensure robustness, bootstrap confidence intervals with 1000 iterations were calculated, with XGBoost demonstrating the narrowest interval range (32.942 to 50.967) compared to Random Forest and Gradient Boosting. The findings offer valuable insights into the role of predictive modeling for subscription-based businesses, providing practical tools to enhance growth forecasting and strategic decision-making.