Instagram is a major platform for digital communication and brand promotion, yet achieving consistent post reach remains challenging due to limited analytical insights. This study proposes a data-driven machine learning approach to predict Instagram reach by analysing early engagement signals such as likes, comments, saves, shares, and profile visits. Multiple regression-based models are evaluated using standard performance metrics, and feature importance analysis is conducted to identify key factors influencing reach. The results highlight that combined engagement indicators significantly impact content visibility, enabling more effective and predictive Instagram content strategies.