To date, the vast majority of work on metal AM has been framed in terms of the need to tune processing conditions in order for a specific conventional alloy, such as stainless steel, Ni-based super alloys, aluminum alloys, etc. This approach often overlooks the fact that historically, every engineering alloy has always been designed having a specific synthesis/processing route in mind. In this talk, I will present some recent work trying to address some of the challenges associated with the design of feedstock for AM. In order to design anything, one must have first a design metric and I thus revisit the concept of an alloy "printability". I then proceed to discuss what types of intrinsic (and extrinsic) materials attributes are amenable to optimization for printability. I finalize the talk by providing a couple of examples of 'co-design', in which alloy performance and printability have been taken into account.