The What-If-Not strategy as proposed by Brown & Walter (1969) is one of the most effective strategies for problem posing. However, it has focused only on the aspect of algorithms for generating problems. The aim of this strategy and how it is used to accomplish the aim of the challenging phase are not clear. We need to clarify the aim of the What-If-Not strategy and to establish the process of the strategy for accomplishing the aim. The purpose of this article is to offer a new What-If-Not strategy by clarifying the aim of the challenging phase.