I consider product and process R\&D from the perspective of consumer preferences. In doing so, I find a novel distinction between the two. I find that the value of a process innovation depends only on quantity sold while that of a product innovation depends also on how much buyers are willing to pay for it and hence also on who buys the product. Over time, increased market penetration leads to a product being sold increasingly to people with a lower willingness to pay for quality improvements and hence relatively more of process R\&D - a pattern observed in several industries. I also consider issues regarding social optimality of the division between product and process R\&D chosen by firms, explain why US firms do relatively more of product R\&D than Japanese firms as observed and several intra-industry observations about firm size and R\&D.