How (not) to use simplified models to search for dark matter at the LHC

I will discuss the motivation, the advantages and the problems of using simplified models as a tool to interpret LHC searches for dark matter. I will present a few examples for how this approach can be used to understand the complementarity of different dark matter search strategies. Finally, I will focus on various consistency conditions that should be imposed even on the most simplified models. These conditions can imply the presence of additional new particles and interactions that may change the phenomenology of the model in important ways