A buyer seeks to procure a good characterized by its price and its quality from suppliers who have private information about their cost structure (fixed cost + marginal cost of providing quality). We solve for the optimal buying procedure, i.e. the procedure that maximizes the buyer's expected utility. We then use the optimal procedure as a theoretical and numerical benchmark to study practical and simple buying procedures such as scoring auctions and negotiation. Specifically, we derive the restrictions that these simpler procedures place on allocations and compare them with the optimal allocations to generate insights about the properties of these simpler procedures and identify environments where they are likely to do well. We also use the optimal procedure benchmark to compare the performance of these procedures numerically. We find that scoring auctions are able to extract a good proportion of the surplus from being a strategic buyer, that is, the difference between the expected revenue from the optimal mechanism and the efficient auction. Sequential procedures (to which many negotiation processes belong) do less well, and, in fact, often worse than simply holding an efficient auction. In each case, we identify the underlying reason for these results.
