Feedback and Competition in Procurement e-Auctions

We use a lab experiment to examine the effect of feedback on bidder behavior in procurement e-auctions. We compare ‘Rank-only’ to ‘show lead bid’ feedback, two regimes applied frequently by procurement professionals. A choice among the two is often based on rules-of-thumb that rely on initial ‘bid compression’, i.e. the spread of the bids submitted pre-auction. The use of such criteria finds no support in existing economic theory. A common assumption in theoretical auction models is that bidders face no opportunity cost from participating in a dynamic auction. This may not hold in situations where the expected value of a contract does not justify a long-time commitment to the bidding process on the part of bidders. In our experiment participants face the choice of remaining active in an auction vs. exiting and being rewarded with a diminishing outside option. Showing the lead bid accelerated bidders’ learning. In the presence of opportunity costs, this can lead to substantially different outcomes conditional on the initial bid compression. With low bid compression, the bidder with the lowest cost wins more frequently, enhancing efficiency, but faces reduced competition by the others, which hurts the buyer’s potential outcome. The opposite is true when bid compression is high. Rank only feedback achieved similar overall levels of efficiency, with higher benefits for the buyer. Crucially, these outcomes are not as sensitive to initial bid compression as in the case of ‘show lead bid’ feedback. A discouragement effect emerging in the ‘Show-lead-bid’, but not in the ‘Rank-only’ regime can explain these results.