Currently, interpersonal trust in computer-mediated communication is a research topic for e-commerce as well as usable security researchers. While the e-commerce researchers focus on gaining warranted trust, usable security researchers focus on preventing misplaced trust, in order to protect users from social engineering attacks. In this paper an approach to integrate findings and theories from both fields is proposed in order to create a complete model for predicting trust in electronic messages or websites, whether they are authentic or not.