| Indonesia has suitable geographic for planting, various types of good quality commodities produced from Indonesia. Coffee is one of Indonesia's export commodities which are well known in foreign countries. One of good quality coffee producer is a province of North Sumatra. The past years coffee exports volume and value has up and down in Indonesia when coffee accounted for sizeable foreign exchange for Indonesia so research on the coffee supply chain in North Sumatra is necessary. Actors identified are farmers, traders, exporters, and government. The flow of the coffee supply chain and the actors who played from coffee plantation to be exported the risks can be identified using ANP method with qualitative data based on interviews. Then, to determine the optimal outcome of income and production for farmers it uses Goal rogramming. ANP method is a development of the AHP to determine the relationship between the factors and actors in the research along with feedback to influence the actor for that factor. Goal Programming is a mathematical model of the development of the linear model that aims to optimize a problem with using variables and deviations for troubleshooting. The results of this study stated that farmers are the greatest actors in the supply chain and owner of the biggest risks of 0.52961 followed by exporters, traders, and government. The biggest risk in this coffee supply chain is price, government support, productions, and quality respectively. According to previous studies known farmers'income in a year and the cost of production. So along with the results of the risk weight ANP, processing the data by Goal Programming to generate optimum value for maximizing revenue and production of farmers with costs of Rp 53,853,790 and fertilizer costs of Rp 3,401,481. The deviation results showed that the condition of the target has been reached with the current conditions. Data used in this research only valid for one year for the costs and farmers income. |