Models have proven themselves to be the key catalyst of many new ideas in science. However, it is not yet fully clarified why models can fulfill such an important heuristic role. The two main reasons stated in the literature—the mental simulation of various scenarios and the wide cross-fertilization across various disciplines—seem to leave out one of the most obvious features of models: they are designed for a purpose. Therefore I investigated why, while the construction of models is a goal-oriented task with a predefined purpose, the use of models yields so many new ideas in science. This paper presents my conceptual analysis together with a detailed historical case study. The functional design of models forces scientists to explore vigorously older ideas to adapt them: as the lacunas in a functional model are also functional, scientists need to modify older ideas (that were formulated for different purposes) to fit the present functional gaps in their models. As such, they construct new ideas. The detailed historical case study exemplifies this by showing how Pauli's original suggestion of the neutrino was, in fact, such an adaptation of Rutherford's earlier idea of the neutron. The present analysis and case study suggest that functional adaptations are salient but often overlooked features of model based investigation.