Understanding and utilizing the full range of built-in spreadsheet functions remains a significant challenge for end-user programmers, leading to inefficiencies, and underutilization of powerful features. This paper addresses this gap by introducing a novel recommendation system that leverages collaborative filtering to intelligently suggest relevant spreadsheet functions based on user behavior. By analyzing an extensive dataset of spreadsheet interactions, this study uncovers user preferences and evaluates the effectiveness of two collaborative filtering approaches against a baseline model. Results indicate that the item-based collaborative filtering method outperforms other techniques, delivering the most accurate recommendations with minimal suggestions, thereby improving both user efficiency and feature adoption. Future research will extend these insights by incorporating diverse spreadsheet datasets and conducting real-world usability assessments to further enhance recommendation accuracy and user experience.