<![CDATA[Indonesia's non-cash retail payment channels exhibit markedly heterogeneous growth patterns, yet forecasting studies in the payment systems domain have rarely compared multiple time series models at the individual channel level or connected forecasting outputs to a management framework capable of informing differentiated strategy. This study addresses both gaps by comparing the forecasting performance of ARIMA, SARIMA, and Prophet across five non-cash retail payment channels in Indonesia using monthly transaction volume data from January 2015 to February 2026, and by classifying the resulting five-year projections according to product life cycle theory. Model accuracy was evaluated using mean absolute error, root mean squared error, and mean absolute percentage error on a holdout test period. The results show that no single model performs best across all channels: ARIMA provides the most accurate projections for two channels, Prophet performs best for two channels, and SARIMA performs best for the remaining channel. Based on projected growth through 2030, three channels are classified as being in a growth stage and two in a maturity stage, each warranting a distinct strategic emphasis. These findings offer a replicable approach for translating channel-level forecasts into differentiated, theoretically grounded payment system development strategy.]]>
Copyrights © 2026
