Network planning tools are routinely used by operators to improve the system performance and to provide a satisfactory service with minimal deployment expenditure. In this context, the deployment flexibility of relay nodes, which mainly stems from the wireless backhaul link, compact physical characteristics, and low power consumption, can be exploited to enhance the system performance. In this paper, we model and analyze two simple approaches that can be used for the planning of two-hop cellular relay networks. The aim of the so-called relay location selection and serving cell selection approaches is to enhance system performance by improving the quality of the wireless relay backhaul link. Location selection takes into account the shadowing properties at the different possible relay node deployment locations, considering only the link quality toward the serving base station. On the other hand, cell selection considers the case where a relay node performs cell reselection from a severely shadowed serving base station to a neighboring cell that is less shadowed. A simple model for evaluating and analyzing the impact of both network planning techniques on the system performance of relay deployments is given. In addition, we present the analytical framework through closed-form derivations of the signal-to-interference ratio (SIR) and end-to-end rate distributions. Performance evaluations that investigate the impact of site planning on the quality of the relay backhaul link, end-to-end rate, resource allocation on the two hops, upper bounds on planning gains, access-link limitations, and the deployment of multiple relay nodes are carried out. Results show significant improvements, which justify the need for relay site planning in relay enhanced networks.