Sponsored Data: A Game-Theoretic Model With Content Provider Content Quality Differentiation

Abstract

Excessive traffic consumption creates anxiety about traffic costs and encourages the popularity of data sponsorship, a business model in which internet service providers (ISPs) encourage content providers (CPs) to subsidize consumers’ mobile traffic costs. In practice, content with data sponsorship may be output at higher or lower resolution. We propose a game-theoretic model in which three cooperation options exist between the ISP and CP: Case N (no data subsidization is allowed), Case L (allowing the CP to subsidize low-resolution content), and Case H (allowing the CP to subsidize high-resolution content). We find that the ISP chooses Case H when the ad-revenue rate and degree of increased viewing cost for low-resolution content compared with high-resolution content (DIC) and degree of increased traffic for high-resolution content compared with low-resolution content (DIT) are high. However, the ISP chooses Case L when DIC and DIT are low and Case N when the ad-revenue rate is low. The CP offers full subsidization to cover consumers’ traffic costs under Case L but only partially subsidizes data under Case H. In addition, the Pareto zone shows that a large ad-revenue rate and a low DIC allow Case L to benefit both the ISP and CP, but a large DIC can let Case H benefit both parties, which sheds light on the motivation behind ISP–CP cooperation from a new perspective. We further identify the conditions under which consumer surplus and social welfare can benefit from a data plan.