Indonesia as the 4th largest coffee producing country in the world escorts companies that process coffee raw materials into roasted coffee to be remarketed to partners at home or abroad. To maintain the company's reputation to its partners, coffee processing companies must maintain the quality of their products or services to partners. The quality of the coffee products needed is specific. The fluctuating values of coffee production supplied over the last 3 years makes it difficult for coffee processing companies to predict the amount of coffee raw materials needed. The use of Single Exponential Smoothing method to predict the amount of coffee raw material stock provides a useful tool for companies to plan and manage their inventory. The characteristics of the data used in this study are univariant for the last 3 years which have very varied characteristics, so the focus of this research is the single parameter alpha. Based on trials and validation using several alpha weights, for the case of data with fluctuating values, it is most suitable to use alpha weights of 0.9 with stock predictions that must be provided by coffee processing companies as much as 180,270 kg of coffee