Geographical indication is an essential label for industrial products. Herein, we aimed to explore a method for establishing geographical indications based on microbial diversity data. We collected two groups of datasets available on the public server of the European Nucleotide Archive. These datasets contain 12 (twelve) NGS-generated reads (amplicon sequencing metagenomes) of fermented cacao beans from Brazil and Mexico. We extracted the microbiome profile using bioinformatic tools in the SHAMAN server. We analyzed further using Principal Component Analysis, Clustering (Ward’s Method of Hierarchical Agglomerative Clustering), and UMAP (Uniform Manifold Approximation and Projection) combined with KNN (K-Nearest Neighbor). We discovered differences in microbial diversity and unique taxa in the fermented cacao beans from Brazil and Mexico. Lactic acid bacteria (LAB), such as Liquorilactobacillus, Tatumella, Leuconostoc, Companilactobacillus, and Limosilactobacillus, are unique genera in samples from Mexico, while Bacillus is a unique genus found in samples from Brazil. We have demonstrated the separation of the microbiome profiles between fermented cacao beans from Brazil and Mexico using PCA, clustering analysis and UMAP-KNN. We have successfully developed the proof of concept in establishing geographical indicators based on microbial diversity data or microbiome profiles. In the future, we will extend this research to analyze samples from Indonesia and establish a microbial diversity database of Indonesian fermented cacao. This database is essential for the authentication assay of Indonesian fermented cacao and for developing fine cacao and specialty products.