Abstract
Object. Demand forecasting is a vital step for optimal organization resource planning. Many factors may influence the future volume and turnover of railway freight transportation. The quality of forecasts made by expert methods no longer meet the requirements of the time. The purpose of this article is to compare the accuracy of the forecasts by two different methods with the fact. Methods. This article uses theoretical research methods, such as auto regressive integrated moving average (ARIMA) and expert methods for forecasting and comparing the results using the percentage mean absolute error (MAPE) and mean absolute error (MAE). Findings. The results of the comparison show great promise for using time series analysis to improve the quality of the demand forecasts for railway freight transportation in Kazakhstan. Conclusions. Time series analysis can be introduced into the practice of the largest enterprises in Kazakhstan, including in the transportation industry. These techniques can significantly improve the efficiency of enterprises through better planning of their operations.