Linear and circular consecutive $k$-out-of-$n$ systems are very popular models in reliability theory, survival analysis, and biological disciplines and other related lifetime sciences. In these theories, the failure rate function is a key notion for measuring the ageing process. In this paper we obtain some mixture representations for consecutive systems and we apply a mixture-based failure rate analysis for both linear and circular consecutive systems. In particular, we analyze the limiting behavior of the system failure rate when the time increases and we obtain some ordering properties. We first consider the popular case of systems with components having independent and identically distributed lifetimes. In practice, these assumptions may fail. So we also study the case of independent non-identically distributed component lifetimes. This case has special interest when a cold-standby redundancy is used for some components. In this sense, we analyze where to place the best components in the systems. Even more, we also study systems with dependent components by assuming that their lifetimes are exchangeable.