Posted
Nov 2007
Place Your Bids
Strategies for selling and buying in auctions
Based on the Research of Gillian Ku, Adam D. Galinsky And J. Keith Murnighan
Common sense might make you cringe at a starting bid of a dollar in an auction for your mint-condition car or a penny for your never-worn designer suit. Conventional wisdom says that if you ask for a high price you should receive a high price. But this adage of “ask and you shall receive” might not always hold true. In online auctions such as eBay, you might even be selling yourself short by asking too much.
The traditional belief is that a high first offer results in a high selling price. This belief is rooted in a psychological principle known as anchoring, such that the size of an opening bid is a testament to the item’s value. The numerical value of an anchor sets the tone of more or less prestige of the product up for auction. Furthermore, a high opening bid can spark the rose-colored glasses effect, leading bidders to focus on the more positive features of an item.
Recent research by Adam Galinsky and J. Keith Murnighan, professors in the Management and Organizations Department of the Kellogg School of Management, and Gillian Ku of London Business School turns tradition on end. In a paper published in the Journal of Personality and Social Psychology, the researchers show that low rather than high opening bids—for a variety of products from shirts to fancy rugs in online auctions—generate high selling prices, demonstrating a reversal of the anchoring effect.
High Traffic, High Prices
Using data collected in laboratory studies and through the eBay archives, the researchers performed a series of experiments to investigate the effect of low opening prices in generating high selling prices and the psychological forces involved in this reversal. In an experiment with MBA students, more participants said they would bid in an auction with a starting price of $1 compared to a starting price of $10. The same pattern held true when participants were asked to imagine bidding on a shirt with an opening price of $1 or $24.99, since the lower starting price spurred more traffic.
However, would the increased flow of auction traffic actually lead to a higher selling price? Could it be that temptingly low starting prices increase the number of bidders and drive up the final price? The researchers scrutinized data from actual eBay auctions to see whether the same product offered at different opening bids would command different selling bids.
To allow the results to be generalized across goods, the researchers chose two different products—a Nikon camera and a Persian rug—to compare the relationship between the initial asking and the final selling prices. “Nikon cameras are plentiful, and people generally know how much they’re worth whereas the Persian rugs are one of a kind and hard to price,” Galinsky said. Data including the starting and ending prices and the numbers of bids and bidders were analyzed from auctions for 179 Tabriz Persian rugs and 87 Nikon digital cameras. For both products, the researchers found that a low starting price led to a higher final price. They also verified that lower starting prices led to more bids and more unique bidders, indicating that traffic contributed to escalating the final price.
The Power of Commitment
Probing in more depth, the researchers wondered what psychological factors contributed to the bidding fervor produced by low starting prices. Perhaps low starting prices engaged people and made them more committed, leading them to place more bids.
To test for this escalation of commitment, the researchers correlated the time invested in an auction with the final cost in eighty-nine online auctions for Tommy Bahama shirts, high-end Hawaiian-themed silk shirts for men that sell for about $100 retail.
When the researchers compared two different starting prices—$9 .99 and $24.99—they found that the more time bidders sank into the auction the greater the final selling price. “Because you lower barriers to an auction, you get them to invest time and resources,” Galinsky said. “They get trapped.”
Following the Crowd
In addition to the time invested as a psychological factor promoting bidding, the researchers wondered whether high traffic attracted more bidders and increased the perceived value of the item. Just as you might be tempted to eat at a restaurant with a car-filled parking lot, perhaps online bidders are drawn to an auction with many bidders.
In an experiment involving an imaginary auction for a Cancun vacation, undergraduate participants indicated how much they thought the trip was worth after seeing an eBay-like page showing the number of bids and the current bid at the pretend auction. With the number of bids ranging from five to twenty-four and the current price ranging from under $900 to about $1,600, participants who saw more bids thought that the trip was worth more than participants who saw fewer bids. “People use the number of bidders to infer value. They think, ‘They must know something that I don’t,’” said Galinsky of this herd effect at auctions. “Economically you want to avoid high-traffic auctions, but psychologically you are lured by high traffic,” he added.
The strength of the study was the consistency of a low starting price leading to a high final price for all kinds of products. “It was true for three really different kinds of items,” said Murnighan of the shirt, camera, and rug. “For the phenomenon to occur with all three gives me a lot of confidence. Traffic is really the key.”
Strategies for Selling and Buying in Online Auctions
To further probe how the buzz at an auction boosts the final price, the researchers studied how decreased traffic for a desirable product affects the sale. “Once you understand a phenomenon, you should be able to turn it on and turn it off,” said Murnighan. Producing low traffic at an ordinarily well-trafficked auction is one way to turn off the phenomenon.
Misspellings of product names on eBay can decrease traffic by keeping potential bidders from finding the auction. Murnighan, an eBay user with a feedback score of 870 and an approval rating of 100 percent, said that misspellings of products are frequent on the Web site and are well known among aficionados such as himself. “You can find ‘Armoni’ instead of ‘Armani,’” he said as an example.
The researchers used data from 43 correctly spelled Michael Jordan and 33 incorrectly spelled Michael Jordan auctions for shirts to see how the starting and ending prices compared with the bid history. If high traffic is the key for auctions that start low and end high, then the pattern should not hold true in auctions with less traffic. That is, one should see a positive correlation between starting and ending prices for misspelled shirts, driven by decreased traffic.
That is exactly what the researchers found: the low traffic of incorrectly spelled auctions meant that low starting prices led to low ending prices. Accordingly, getting a bargain depends on finding a low-trafficked auction. Look for misspelled items, items with no photographs, or items with auction end times outside the bid-frenzied prime-time hours. “You want to find a market that is as narrow as possible,” Galinsky said. The reverse is true for sellers. “Sellers should create a market that is as wide as possible,” Galinsky said. “You start with a low price when you believe that there’s a substantial market for the item such that you lower the barriers for people to start bidding.”
9 Comments
Nov 8 2007
for the cancun example: could simple economics explain this perhaps?
The supply line is flat (1 cancun vacation is being auctioned)
For the demand line, we have 2 cases. Lets say…
Case 1: 100 bidders
Case 2: 50 bidders
Both of these “Demand” lines intersect the same supply line but since in Case 1, the demand # is higher, its intersection is to the right of Case 1 demand line which means it has a lower incline (gentler slope) moving from top left to bottom right. So we now have 2 demand lines with differing slopes. THe problem is in an auction, the supply side was fixed but inherently one would think that the supply line was also a regular supply line and icnreased with price, then the Case2 line would now intersect at a higher point - in other words left to free market, the Case2 was worth more to the market.
Very compelling points about setting a low barrier to entry and participants then getting locked in thanks to time (and ego?) commitments.
Nov 8 2007
Great idea to analyze our cancun study with simple supply and demand.
the one factor that’s not a part of this analysis, however, is that the demand varies: some people are willing to spend more, and some less. this value distribution is not likely to change much with more bidders - yes, there are chances for more bidders at the tails of the distribution, but this should only increase the final bid marginally.
Instead, we see a marked increase as a function of the number of bidders. Thus, it’s probably more than simple supply and demand - but that, too, may be a part of the effect.
Nov 10 2007
Just coming back from an ebay auction that I won. This article is great and let’s me improve my buyer strategy. Since traffic for product auctions in most cases is correlating with the auction time to end, i.e. the sooner the auction’s end the higher the traffic. It is beneficial from a buyer’s point to very early in the game lift prices considerably on interesting products to a level of 50-60% of the expected end price. With much traded products an approximate end price is possible to estimate. This way traffic is slowed down.
Then, further not engage in the auction at all but rather set a timer alarm on your cell for 3 minutes before auction ending time. If the price of the auction is still interesting, you prepare for an “attack” in the last 15 seconds by entering a bid that’s 5 - 10% higher than the currently highest bid. Stick with odd numbers in the end, e.g. 110,12 EUR or so. This behavior is much seen on the German Ebay site. One has to know one’s max price of course, but it’s easy to keep a cool head since the decision to really engage in the auction comes when the current price is close to it’s ending price.
The definition of an ending time in Ebay is the key point to the latter strategy. Other auction platforms (such as huuto.net) in Finland extend the time of an auction by 5 minutes if a new highest bid is entered, that gives others the possibility to react and renders “last second attacks” useless.
Nov 27 2007
Thanks for sharing your research.
Prof. G. taught us this in class and I have implemented it whenever possible. However, I did find that it is very difficult to choose an anchor, especially when there are multiple variables in the discussion and you don’t know what the other person is feeling (they may get offended). For example, hiring discussions on compensation. In fact, I would argue that there is a meaningful segment of the population that should not follow the advice of this research simply because they do not have good judgment or lack self-control to open with an appropriate anchor. Just a thought since I find that I am one of these people who seems to get very greedy when opening with the first offer. I see my mental flaws and, now in some cases, I let the other party anchor first. But I ‘know’ they are anchoring and I just counter anchor and wear them down.
Nov 28 2007
Thanks for your comment
Although I do subscribe to the belief that going first is still on average in your best interests. The key is to have an ambitious first offer mapped out in advance. Your first offer should be based on your target price. So should your first counter offer. That is why I always tell people to write down their first offer in advance. That way, they can’t be anchored on their opponent’s first offer. That said, when you go first, you will anchor your opponent’s counter offer as many of mine and other studies have demonstrated. So going first is still a good idea!
Great to hear from you!
Feb 6 2008
I find this researching fascinating. As a Kellogg alum (1990) who started her own eBay dropoff store 4 years ago, I must say that the research supports my experience.
Whenever we have set price highs on eBay, it was when we started low. That doesn’t mean you don’t research your item first, including key title words and description - that is critical - it just means that you know its approximate value and the optimal way to realize it is through a low opening bid.
The most difficult aspect we have as an eBay store is dealing with our client’s expectations. If there have been any updates to this research or comments, I would be happy to learn about them. Susan
May 13 2008
During the most recent real estate boom, many buyers were outbidding each other. Does your theory work in the real estate market also?
May 14 2008
A great question. Soon after our research came out, we saw an article in a local paper citing our work, in conjunction with a house that was being listed with a starting price of $1! We weren’t able to follow up on that example, so I don’t know what happened to it, but the principle is simple: if, by reducing barriers to entry, you can entice multiple bidders, the final price is likely to be high. Our caution in the real estate arena comes because of the high stakes: if you put a low price on a house and you don’t generate much interest, you may have done yourself a serious financial disservice. At the same time, we’d love to be able to buy a couple hundred houses, set the price low on half of them and higher on the rest and see what the data tells us. It would be very illuminating! The best test would be if we could buy 100 pairs of identical houses - and set the price very low on one and a higher price on the other. I would bet that our findings would apply - but since we’ve only studied this with lower stakes objects (cameras, oriential rugs, etc.), it’s still a bet.
Jul 27 2009
Here’s the problem: eBay is not a true auction site. eBay creates an artificial barrier with their incremental bid restraints, deliberately preventing buyers from submitting an actual dollar amount, (thereby preventing sellers from obtaining the highest and best dollar amount their auction item could have commanded). To wit; If you have bid one dollar on an item and I as the second bidder am willing to top you by bidding one hundred dollars for the item, the actual bid reflected on eBay will only reflect a bid for TWO dollars instead of the one hundred dollars I actually bid. The price dampening psychology works well for eBay, by displaying exceptionally low prices right across the board, which then generates a larger audience. The seller, however, is most definitely getting the short end of the stick from eBay’s behind the scenes price rigging. More importantly however, what happens as a result of this artificial manipulation of the percieved market, is that the ebay audience, and to be sure, it is a very large one, begins to think the artificially manipulated price range they see displayed is the actual market value of these items, which has the effect of depressing the national market for these same goods. eBay itself does not incur a financial loss by artificially depressing the market because their final value fees have now been increased to the point where eBay now takes up to a whopping 10% of the sales price of lower priced items, in addition to taking their bread and butter listing fee charges off the top. It would be interesting to see a test market section erected on eBay which would permit a study of bidder behaviour within a true auction format.