Behavioral research and empirical modeling of marketing channels: Implications for both fields and a call for future research