This paper analyzes the bahavior of a principal with bounded memory who can offer a two-period performance-based contract to an agent. In the model he can choose whether to evaluate the agent after each period or only at the end of the second period. If the agent is wealth-constrained, the option to evaluate him twice can be profitable. But without the constraint on the part of the agent, the principal will always prefer to evaluate only once in order to reduce evaluation costs. Finally we consider a modification leads us to a trade-off between more and less frequent evaluations.