In this paper we used natural language processing to review hundreds of thousands of negative student reviews of their teachers submitted to the website RateMyTeacher.com. Our analysis identified several issues raised by students when rating teachers poorly, which adds to the literature that defines "bad teachers" from the student perspective. We also identify the language students used to describe these issues and notice a clear distinction between the language used to address teaching-related complaints and behavior perceived as unfit for teachers. We argue that digital forums can be valuable tools for schools and conclude with suggestions and examples of the type of policies that may be derived from this type of analysis of a digital forum for students.