Split-Award Auctions for Supplier Retention

Abstract

T o stay abreast of current supply-market pricing, it is common for procurement managers to frequently organize auctions among a pool of qualified suppliers (the supply base). Sole awards can alienate losing suppliers and cause them to defect from the supply base. To maintain the supply base and thereby control the high costs of finding and qualifying new suppliers, buyers often employ split awards, which in turn inflate purchase costs. This results in a trade-off that we investigate in an infinite-horizon stationary setting in which the relative cost position of each supplier is randomly drawn in every period. We characterize the optimal split award that minimizes long-run costs (purchasing and qualification) and show that maintaining a constant supply-base size-using a "qualify-up-to" policy-is optimal for the buyer. We find that neither the extent of multisourcing nor the buyer's value of split awards compared with winner-take-all auctions are monotonic in the qualification cost and that split-award auctions can increase ex ante system, buyer, and supplier benefits simultaneously. To our knowledge, this is the first paper studying split-award auctions for supply-base maintenance, and it will hopefully galvanize further research on this important topic.

Supply Base Design for the Procurement of Multiple Items

“…Our model also reveals some limitations: for simplicity, we assumed that the supply base is fixed during the planning horizon and only allocation to suppliers varies over time. In reality, supply base management is a continuous process where each year, some suppliers are lost and new suppliers are qualified (Chaturvedi et al 2014). Future research should provide a closer look at these dynamics.…”

Split-Award Auctions: Insights from Theory and Experiments

“…Equation (8) implies that if the buyer gave Q n amount of business to each of the n bidders at pricec and then gave Q n 1…”

“…In the two non-regular distribution treatments, equilibrium bids for costs above 25 jumps to very close to 100, so while the equilibrium bid function for the cost range of 0-25 can be reasonably approximated by a linear function, the equilibrium bid function for the entire 0-100 range is clearly non-linear 8. Although optimal functions are non-linear (with the exception of the 100-0-0 split) non-linearities are not very strong, and the linear approximation captures the e↵ect of the intercept and the cost, providing a fair comparison for the estimates.…”

Supplier Competition and Cost Improvement

“…This stream of literature has considered a range of issues that affect supply base decisions, including supplier qualification costs (Riordan 1996, Chaturvedi et al 2014, supplier diversification (Babich et al 2007, Wan and, production diseconomy (Tunca and Wu 2009), supplier learning (Klotz andChatterjee 1995, Lewis andYildirim 2002), up-front concession (Li and Debo 2009), fixed supplier costs (Agrawal and Nahmias 1997), and investment in supplier reliability (Wang et al 2010). We contribute by examining the impact of supplier competition on suppliers' incentive to improve, which we show can be positive or negative depending on the information structure.…”