Complex societies are suggested to generate complex communication. However, this has rarely been empirically tested. For a systematic approach, we first have to define what we mean by complexity. What defines social complexity, and what defines communicative complexity? What aspects of social complexity correlate with what aspects of communicative complexity? I review the hypotheses put forward for the evolution of the diverse communicative repertoires and variation within mammals, focusing on vocal communication. Within the social domain, I will examine the influence of group size, group composition and group structure, as well as social interactions differentiated into cooperative and competitive contexts. Taking the example of communication in meerkats (Suricata suricatta) in comparison to other closely related species with different social structure within the family of mongooses, Herpestidae, I will disentangle the different hypotheses, and compare the findings to other mammalian orders. I argue, only by identifying correlations of communicative variation with specific social contexts and taking the ecology of a species into account, will we be able to understand the evolution of the diversity of signals in different mammal species. Ultimately the function of signals has to be considered from both the producer and receiver side, in the natural habitat of a species, to explain the variation, complexity, as well as potential limitations in animal communication.