Reward programs offer rewards, and with further progression toward these rewards, customers become more motivated. This acceleration is called a reward gradient. This study proposes prospect theory as the underlying mechanism of the reward gradient and suggests a nonlinear relationship between motivation and progress level (low vs. middle vs. high). A numerical simulation based on the mathematical model and experiments were conducted to see how the expected values of rewards can change as progress is made, and how this further affects motivation. The results identified a stronger acceleration in the expected value after reaching a middle point of the reward achievement, which is due to a larger deviation in the loss value than the gain value. These findings can help restaurants and tourism companies implement a dynamic perspective in reward programs for enhancing customers’ reward gradient behaviors.