Reading the Tea Leaves in Online Reverse AuctionsPublished: January 21, 2011 in Knowledge@Emory
Reverse auctions can leave sellers with more than just a bad taste in the mouth. There can be cursing, despair, and a feverish sense that all control has been lost as suppliers competing in these procurement events offer deeper and deeper discounts to compete for contracts. Buyers, on the other hand, often savor the experience. Some order in food just for the occasion, so they can snack while watching how the auction takes bigger and bigger bites out of supplier profits.
But new research suggests that sellers don’t just stew in their juices as online reverse auctions unfold. In some cases, suppliers seem to actively monitor their competitors’ bids to discern who their true competition might be, says Sandy Jap, a professor of marketing at Emory University’s Goizueta Business School. The evidence comes in a recent study of bidding patterns by Jap and colleague Ernan Haruvy of the University of Texas at Dallas. Some suppliers in online reverse auctions seem to regularly bid above the lowest observed bid over the course of an auction, the researchers say. Other suppliers, meanwhile, seem to always respond aggressively to the lowest bid on the table by offering an even lower price.
The researchers argue that the differences likely stem from quality differences among suppliers who, in turn, employ different bidding strategies to compete for contracts. In most online reverse auctions, the lowest bidder isn’t automatically assured of winning the job. The duo’s research suggests that high-quality suppliers carefully monitor auction bids from competitors to make sure they offer some degree of discounts without giving away the farm.
“In some product categories, there might be only three or four or five qualified, viable suppliers out there,” Jap says. “So, when you show up to the auction and all of a sudden you see 10 bidders, suppliers know there are probably some bidders that are just meant to shill. They don’t want to fall into the trap of dropping prices to compete against them.”
Online reverse auctions began to emerge as a new way for companies to procure goods and services in the late 1990s. The events dramatically condensed the amount of time it took buyers to secure contracts while also promising significant price discounts along the way. Today, most large companies own software that allows them to conduct online reverse auctions, in which groups of pre-qualified suppliers are invited to bid for procurement contracts. Reverse auctions now are spreading beyond the Fortune 500 in novel ways, whether it’s the federal government promoting auctions as a way to buy office suppliesor states using the events to award energy efficiency grants. But even with their popularity, online reverse auctions have been controversial. Some argue that the events can damage long-standing relationships between purchasers and suppliers, with one criticarguing that “reverse auctions are simply technology-enabled zero-sum power-based bargaining and thus a worst practice.”
Online reverse auctions typically last about two to four hours and differ from consumer auctions in key ways. For starters, prices in reverse auctions generally go down—not up. More importantly, discounts aren’t offered in a uniform manner. Whereas bidders in consumer auctions usually must top the highest current bid to stay in the game, suppliers in online reverse auction don’t have to go below the lowest current bid. For example, a firm that might offer to supply industrial parts at a price of $90 per part at the outset of an auction might watch other firms aggressively bid the price down to a range of between $60 and $70. Even so, that first supplier might respond by dropping its bid by just five dollars to $85. This isn’t just a hypothetical possibility; Jap says that suppliers often seem to divide themselves into groups of higher- and lower-price bidders.
To try and explain why this happens, the researchers analyzed point-by-point bid data from 10 online reverse auctions sponsored by an industrial buyer in the automotive industry. The combined contract value of the auctions was $19.2 million, and none of the products being sold—metal parts and plastics—was purely a commodity. That’s an important distinction, Jap says, because in the case of such products non-price characteristics play a bigger part in the buyer’s ultimate decision about which supplier to select. While online reverse auctions put a focus on the price being offered by various suppliers, buyers also have time after an event to determine exactly which of the bidders it wants to source the contract to. Buyers may take as much as six weeks to evaluate the bids by visiting the suppliers to inspect their manufacturing process and quality control systems. So, the auction events studied by the researchers involved products where the quality of the bidder—not just the prices offered—could shape the ultimate outcome.
Among the auctions studied by the researchers, the average auction duration was 101 minutes and the events featured a “soft close,” meaning if a supplier offered a bid late in the game, the auction could be extended to allow time for other suppliers to respond. Researchers charted all the bids made during all the auctions and found that more than two-thirds were concentrated in the final quintile of time, suggesting very aggressive bidding behavior as the auctions closed. But the behavior was not universal among all bidders. A minority of suppliers bid relatively infrequently during the auctions, and placed more than one-third of their bids during the first quintile of the auctions. After testing the waters in this way, these bidders placed only 30 percent of their bids during the final quintile of time—a very different bidding pattern than most suppliers. These bidders also seemed to respond aggressively to bids only from a few other suppliers in the auctions—not all suppliers. And the firms were significantly less likely to try and match the lowest bid offered during an auction.
“We find that 62 percent of suppliers are naïve bidders in that they do not appear to derive information from others’ bids,” the researchers write. “The other 38 percent of participants appear sophisticated in that they attempt to infer others’ qualities from their bids.”
The observed behavior closely matched the results predicted by a model that assumed auction bidders differed in terms of quality. Firms that were consistently aggressive in bidding down prices likely were playing to their strength as “low-quality” competitors, the researchers say. These low-quality suppliers focused their energy on discounts because they likely lacked a history working with the buyer and other non-price advantages such as logistics or specialized personnel. The more nuanced approach to bidding came from “high-quality” suppliers, and this behavior fits with previous research from Jap and Haruvy. In a 2008 paper, the researchers showed that suppliers willing to make investments in a collaborative business relationship with buyers were less willing to discount prices in auctions. The new paper goes a step further in suggesting that high-quality suppliers in online reverse auctions have distinct goals compared with low-quality suppliers.
“Bidders may be following a bidding strategy intended to keep them in the buyer’s (segmented) consideration set for final choice,” the researchers write. “This is a very different strategy than one in which the bidder bids to the minimum closing price they can accept. In fact, the strategy observed in this research would suggest that the bidder need only give the appearance of being competitive with other like bidders in order to remain considered for the final choice.”
The research offers a few take-home messages for buyers in online reverse auctions. First, high bids don’t necessarily signal a lack of interest or aggression on the part of a high-quality bidder; the firms are just operating according to a different perception of the market relative to the perceptions of low-quality bidders. So, buyers should be aware of this “bidder segmentation” when evaluating auction results, the researchers say. What’s more, if buyers are concerned that high-quality bidders are avoiding competition from low-price competitors in the auction, they could change the format of the events. Suppliers in online reverse auctions don’t know the identities of other firms competing for the contract, but bids that appear on computer screens during the auction are linked to unique identifiers for various participants. By letting participants only see bid values, buyers could change the auction format to “take away (suppliers’) ability to learn and infer who their relevant competitors might be,” Jap says.
For sellers in online reverse auctions, the research points out a “silver lining,” Jap says, to one of the most upsetting aspects of the events. Suppliers are very protective of information about the prices they’re willing to charge for their products and services, Jap says. So, online reverse auctions can make bidding firms feel very uncomfortable. But by closely monitoring the bids offered by competitors during an auction, suppliers can be strategic with their bids—and potentially hold at least some of their pricing information closer to the vest.
“From a high-quality bidder’s standpoint, you don’t want to bid to your bottom-line price— you just want to appear competitive relative to everyone else like you in the auction,” adds Jap. “So, the good news for suppliers about a completely open auction is that you can try to figure out who your competition is and bid to their levels; this may mean that you don’t have to give away the farm.”
With more and cleaner data, researchers say they could evaluate whether dynamic strategies such as price signaling and collusion might be at play in online reverse auctions. But the results already in hand help explain why online reverse auctions don’t function according to many of the rules and theories that have been established over the years through research focused on conventional auctions.
“Standard auction theory assumes that all bidders are equal,” Jap says. “The problem is, in a lot of these industrial sourcing auctions, buyers don’t just care about the price—they also care about other things such as the quality of the product and the responsiveness of the supplier. Does the supplier help me in a crisis, or will he try to evade my calls? Is he willing to work with me to help design a delivery system that works for all the plants I have? Is he flexible? Is he trustworthy? In standard auction theory, the buyer is indifferent to which bidder wins—all bidders offer the exact same value.”