Predicting revenue is a crucial concern for defining organizational strategic goals. Business strategies are largely dependent on the status of the product’s revenue. Various statistical and analytical techniques are available to forecast business revenue which may not provide a precise prediction due to considering all relevant predictive indicators or features. This study addresses the critical importance of accurately predicting mobile revenue in a highly dynamic and complex market influenced by diverse factors such as product type, regional 5G coverage, and market dynamics. The primary objective was to create an effective forecasting platform that uses powerful machine learning models to predict the nonlinear nature of relationships that drive mobile revenue generation. The novelty in this work is in the design and implementation of an ensemble strategy called Optimal Weighted Blending (OWB) which combines the best Gradient Boosting Machines (GBMs) such as CatBoost, XGBoost, and LightGBM and gets high predictive accuracy and stability. The study utilized an extensive dataset on a quarter of mobile sales of between 2019 and 2024 that has detailed features such as units sold, market share, and features indicators of the 5G ecosystem to perform rigorous preprocessing and feature engineering to prepare the data and train and evaluate the model. It was found that OWB ensemble was better than individual models with an R2 of 0.91 and a mean absolute error (MAE) of 2.37 million successfully modeled complex interactions among features and mitigated prediction variance. Consequently, the findings of this study would be beneficial to organizations in estimating the revenue of any product, as well as in surviving in a highly competitive world.