Foresight to 2035 for specialty coffees: scenario alignment, the Delphi method, and integration with Python libraries

Purpose: Current trends have positioned the specialty coffee chain (SCC) as an axis of interest within the coffee industry, being a prioritized chain in the department of Quindío, Colombia. Considering the planning and execution needs of government programs and policies and the difficulties for decision-making of public and private organizations in the sector, prospective studies are emerging as fundamental tools for anticipating the future. The purpose of this study is to identify priority technologies and future scenarios for the specialty coffee chain in Quindío, Colombia, based on the participation of key stakeholders and sector experts. Design/methodology/approach: Two prospective methods were implemented, the scenario method and the two-round Delphi method, to prioritize the key variables, technologies, actors, innovations, objectives and scenarios for the SCC up to 2035. Findings: Based on the analysis of 81 variables, 7 key variables were identified and defined by experts through the MICMAC analysis: “Cooperatives, Associations for the process,” “Category Origin,” “Category – Sustainable,” “Production chains,” “Associated crop,” “R&D&I” and “Potential quality”. Based on these, two future objectives and five hypotheses were defined, which served as input to construct the scenarios according to their probability using the matrix of alliances and conflicts: tactics, objectives and recommendations method. The Delphi method addressed 130 topics, classified into 9 thematic categories, and these were rated in the first round and reclassified in the second round according to the criteria of the industry experts who participated in the study. Fifty-eight topics were a priority, 72 were under discussion and 0 were not a priority. Originality/value: The novelty of the study consists of the alignment of methods, techniques and tools for scenarios and expert consultation, as well as the implementation of an artificial intelligence code and natural language processing in Python, which allowed validation of several of the results of the methods used.