As the importance of the effect of customer referral behavior on firms' marketing performance has recently been highlighted, a series of studies have been conducted on how to promote customer referral behavior. However, most research on strategies for promoting customer referral behavior has been primarily focused on reward scheme design, raising the need for developing more diverse incentive approaches. In this study, drawing on the model of Hauser, Simester, and Wernerfelt (1994), which examined salesperson incentives for enhancing customer satisfaction, and the model of Chu and Desai (1995), which examined distributor incentives, and building upon the models of Biyalogorsky, Gerstner, and Libai (2001) and Chen and Shi (2003), which examined reward incentives for customer referral behavior, we investigated whether cost-subsidy incentives can be an effective means of promoting customer referral behavior beyond reward incentives. Furthermore, we examined through a game-theoretic model how customer referral reward incentives and customer referral cost-subsidy incentives change in response to variations in referral costs, and explored their implications. The main findings of this study are as follows: (1) By using both customer referral reward incentives and customer referral cost-subsidy incentives together, both firms and customers can benefit more. (2) When the overall referral cost burden borne by customers is high, it is desirable to consistently reduce both customer referral reward incentives and customer referral cost-subsidy incentives. (3) When the referral cost burden that is difficult for firms to subsidize is high, the effectiveness of customer referral reward incentives is greater, so it is desirable to utilize customer referral reward incentives more actively than customer referral cost-subsidy incentives. This study can be viewed as one of the pioneering studies on cost-subsidy incentive approaches as a means of promoting customer referral behavior. However, it lacks empirical verification, leaving room for complementary research in the future.