<![CDATA[Ecotourism micro, small, and medium-sized enterprises (MSMEs) often face highly volatile demand characterized by frequent zero-sales days, strong seasonality, and exposure to external shocks. In such conditions, sustainability depends less on forecast accuracy and more on timely, low-cost operational decisions. This study examines how forecast-triggered decision-making supports short-run viability under intermittent, zero-heavy demand. Using manually recorded daily sales data from ecotourism MSMEs in Tawangmangu, Indonesia, a two-stage approach is applied that separates sale occurrence from sales magnitude. First, a logistic model estimates the probability of a sale to generate early-warning signals. Second, conditional sales magnitude is predicted to indicate readiness levels rather than precise revenue targets. Instead of focusing on accuracy alone, the analysis evaluates decision usefulness through time-ordered backtesting, emphasizing avoidable operating days and early-warning lead time. The results show that sale-occurrence signals effectively guide daily operating decisions, while magnitude forecasts support proportional readiness. The framework identifies a substantial share of avoidable operating days and provides several days of advance warning before prolonged zero-sales periods. This enables earlier cost control and capacity adjustment. The study contributes by offering a practical, human-in-the-loop decision framework that links demand uncertainty with adaptive actions using simple, manually recorded data.]]>
