Does a space or a group admit a differential graded algebra model with certain desirable finiteness or formality properties? In this talk I will outline several obstructions for this to happen. Some will involve minimal models and Malcev Lie algebras, while others will involve generalized Massey triple products.