Current automated visualization techniques use only a small set of basic guidelines, mostly based on perceptual principles. Yet, in literature a broad set factors are discussed for effective visualization design. In this paper, we analyzed principles and guidelines extracted from academic papers, books, and blogs and examined factors that influence visualization design to aid the design of automated visualization systems. We used grounded theory to code and examine concepts and relationships extracted. We found a variety factors, including domain, semantics, user tasks, insight, and display characteristics and found a variety of interactions between these factors including causal, contextual, and logical relationships. Our findings suggest that automatic approaches need to handle not only broader set of factors but also complex interactions among them. Furthermore, several factors such as domain and semantics are likely to gain even more importance, necessitating flexibility in terms the specification of design know-how as input to automatic visual analytics systems.