A recommender system helps users to select the desired items by analyzing the user's habit of interacting with the system. Recommender system also help the group of users for selecting items due to information overloads. Group Recommender System (GRS) is designed to identify all preferences within a group. An aggregation strategy is needed to accommodate all user preferences in a group. GRS is required in many cases, for example in the book domain, a bookstore recommends a list of books through a display for a group of visitors. We design a GRS for the book domain using Matrix Factorization technique. We utilize three methods to design GRS, such as After Factorization (AF), Before Factorization (BF), and Weighted Before Factorization (WBF). These three approaches were applied to three different group categories, i.e., small groups, medium groups, and large groups. We aim to find the best approach for each group category in this research. The evaluation metrics used are precision and recall in building this GRS. The results of this research indicate that a small group is suitable for using all three approaches, AF methods is the best approach methods for medium groups, and the best approach method for large groups is WBF.