Using research to improve online auction results

Society is facing a huge problem in terms of the amount of waste it generates. The best option for disposal of your unwanted items is to extend their lifespan (if they’re in good enough condition) by giving them away to people you know or selling them to people who want them, rather than chucking them out or donating them to charities. For a lot of people, selling online is one option, which often means online auctions. I’ve seen a lot of people mention that they feel like online auctions aren’t worth the effort (all the taking of photos, the measuring of dimensions, the dealing with very fussy buyers, the listing and selling fees, etc.) but if you’ve got a decent item that you want to get rid of, selling it through an online auction is possibly one of the more responsible, sustainable things to do.

As such, I’d like to encourage people to sell through online auctions by providing some guidelines that might increase the likelihood of a successful auction, courtesy of good old scientific research on the topic.

There’s been a bit of empirical research looking at what factors in an online auction influence things such as potential bidder interest, bidding activity, and most critically, the final sale price. Some of those factors aren’t within the control of sellers, such as how many bids an item receives early on during the auction, or simply how interesting or valuable in general your item is perceived to be by others. However, some factors are within sellers’ control, and I’ve summarised the research findings below, paper by paper, regarding what the potential risks and pay-offs are for a seller if they make particular choices for their auction.

“Individual and social determinants of winning bids in online auctions”

  • This paper found that the number of photos of the item included in the auction indirectly influences the number of bidders and possibly increases the final sale price. So, the more pictures, the more first-day bids, which encourages more bidders overall.
  • The greater the number of first-day bids, the higher the winning bid. I guess this seems unsurprising, if you consider the number of first-day bids to be an indication of the level of interest in the auction. The authors suggest this is due to herd behaviour, since the early bids indicate that the item is popular which in turn makes other people more interested in it. It implies that the item is of value. This isn’t something you can influence as a seller (unless you’re dummy bidding, which you shouldn’t be doing) but it’s something that’ll help you gauge the likelihood of a successful auction along the way.
  • The higher the starting bid price, the higher the winning bid. The other papers I’ll mention in a moment will explain this a bit further, but in the case of this particular paper, it is suggested that that higher starting prices act as indicators of the item’s quality – you might be a bit wary of an item if the price is too cheap, since that might suggest it’s a fake. However, this doesn’t just mean you can always set your auction starting price as high as you like, since if the price is too high, it will discourage bidder traffic and there may end up being no bids at all on the auction. You can use a higher starting price to indicate value, but it still has to be relative to what people are actually willing to pay.
  • Since a greater number of images seems to encourage more first-day bids, which encourages more participation and interest in the auction and seems to result in higher winning bids, it is possibly worth using the maximum number of images permitted (except in the case of very low-cost items). However, this doesn’t mean that more is necessarily more – there might still be an optimal number when it comes to how many images to use. However, this research seems to suggest that you certainly don’t want to skimp on the images for your auctions.

“Behavioral outcomes from online auctions: reserve price, reserve disclosure, and initial bidding influences in the decision process”

  • This paper looks a bit at the role that reserve prices play in online auctions. Some eBay platforms don’t allow for reserves (Australian eBay.com.au and I think also German eBay.de, and possibly others) but fear not – there’s still information to be gleaned from the research even if you can’t impose reserve prices on your auctions.
  • This paper found that a low reserve price results in more overall interest in an auction and a higher final sale price, when compared to a high reserve price (relative to the price of the product).
  • The disclosure of a reserve price also results in more potential bidder interest, which increases the likelihood of a higher final sale price.
  • Given the previous two points – the ideal situation would be a low, disclosed reserve price, e.g. $1. This is also relevant to anyone using an eBay platform that doesn’t allow reserves, since a low, disclosed reserve price is equivalent to a low starting price – they’re both situations in which the interested potential bidders know that the price at which they could obtain the item is potentially quite low, and this stimulates interest in the auction.

“Determinants of Internet Auction Success and Closing Price: An Exploratory Study”

  • This paper found that using a reserve price resulted in a lower probability of auction success, but for auctions that were successful and did have reserves, the final sale price was higher. Again, the authors suggest that having a reserve indicates quality to potential buyers – the item is valuable, therefore the seller doesn’t want it going too cheaply. However, having an unknown reserve price is a bit of a gamble, as potential bidders see this as a decrease in the probability that their bid will be the winning one. Some people employ rules about which auctions to bid in, and as such may decide to entirely avoid auctions with unknown reserves. They may engage in price aversion strategy, meaning they don’t get involved with auctions with reserves as it makes it riskier that they will unexpectedly incur higher expenses or losses than they had anticipated – they don’t have enough information to base their decision on regarding whether bidding is likely to turn out more expensive than they planned, so they avoid the auction completely.
  • This paper also found that a higher starting price resulted in a lower probability of auction success but a higher average closing price for auctions that were successful. Again, price indicates quality. However, a high opening price might scare away potential bidders. The good thing about a low opening price is not only that it gets more people interested in the auction since they’re all considering their potential to land a bargain, but that it also provides more information to the bidders – more people will bid due to the low price, which gives everyone the chance to see how popular the item is and how much people are willing to pay for it.
  • A low opening price can also encourage what is known as “bidding momentum” – the bidders start bidding at the low prices and they get caught up and invested in the auction, possibly even going so far as to imagine their future ownership of the item. This in turn makes them more likely to keep bidding, and possibly bid more than they originally would have. As multiple people become invested in the auction, the try to outbid each other and generate momentum that might push the final sale price higher.
  • This paper also investigated whether auction duration affected the probability of the auction’s success or final sale price, hypothesising that a longer auction might allow more people to become interested, but it doesn’t seem to generally be the case. It may depend on the item.

“Buying, Bidding, Playing, or Competing? Value Assessment and Decision Dynamics In Online Auctions”

  • Elaborating on the idea that a low starting price could result in more interest and a higher sale price, this paper looked at whether a low starting price would encourage the “herd” behaviour mentioned above, or rather an “anchoring” effect. Anchoring would mean that potential bidders would see the initial low price and then judge any price increases relative to that – so if an item was listed for $5 and the bidding gets the current highest bid up to $10, the other potential bidders might then think “Well, it was a bargain at $5 and I was happy with potentially paying that price, but I’m not willing to outbid $10 because that’s twice as much as what it was”. Their evaluation of the price is anchored to the original starting price. This means that bidding momentum might not build up and the herd behaviour might not occur, meaning the final sale price doesn’t end up being particularly high.
  • So if the conditions are right and the right people with the right prices in mind are involved, a low starting price might result in a bidding war on your item and a higher final sale price. However, this paper found that even when there were a lot of bidders due to a low starting price, it was not necessarily sufficient to create momentum or a bidding war.

As you can see, the research doesn’t necessarily offer any clear-cut, obvious strategy to guarantee a maximal final sale price in an online auction, mainly because there are so many other factors that simply can’t be controlled – the nature of the item you’re selling, the number of people interested in the item, their individual bidding strategies, etc. Including plenty of photos and disclosing your reserve, if you have one, seem to be good ideas, but when it comes to the other factors that you do have a say in, most of the possible choices involve a trade-off. You can set your starting price and/or reserve price high to indicate the quality of the item, but then you risk decreased interest in the item and a lower likelihood of a successful auction, although if you do end up with a successful auction, the final price is likely to be higher than it otherwise would have been. Alternatively, you can set your starting price and/or reserve price low to generate interest in the item, but then that could go either way: it could result in herd behaviour where people consider the item to be valuable because lots of other people are interested in it which could result in people feeling invested in the item, competitive bidding, bidding momentum and possibly a higher final sale price, or it could result in anchoring effects and even though you’ve generated a lot of interest with the low starting price, people are still only willing to make still relatively low bids. Or people could be suspicious of the low price and think that it’s an indicator of something dodgy. So yes, it’s a bit of a minefield, but now you know a bit more about how to negotiate the risks and how to make informed choices about factors in your auctions.

4 Thoughts on “Using research to improve online auction results

  1. Very interesting! Me and some friends are going to do an ebay sell next weekend too. So lots of photos is the key…

    • Yeah, not skimping on the photos certainly seems to be a good idea. I could have added a lot of personal opinions to this post but I tried to keep it to just the ~empirical research~ haha.

  2. I stalked a Hermes scarf I wanted for several days, knowing my top price. The auction was slow to start off and then at the last minute there was a flurry of activity that drove me off. That is a slightly different scenario than what you’ve described. My sense is that 20 or so potential buyers had their eye on the scarf…but like me, were holding off until the last minute to keep the price as low as possible.

    • Yep, the research I mentioned by no means covers the entire range of behaviours you might encounter in an online auction – there are all sorts of factors that probably govern the bidding activity, whether it’s early on or more towards the end of the auction.

      One paper did describe that sort of end-of-auction flurry of bidding, but I didn’t write about it since it was only really relevant to explaining buyer behaviour rather than offering anything that sellers could do to improve their chances of a higher sale price. But the authors of that paper did note that not bidding until the very end of the auction was a strategy most commonly used by experienced buyers, the rationale being that if you hold off on bidding, that prevents too much information from being available to other bidders about how popular the item might end up being. So if other potential bidders think that interest in an item isn’t particularly great because no one has bid on it, they might only prepare themselves to pay a reasonably low price, then be totally trounced and outbid at the last moments of the auction. And like you said, people also seem to not bid early because they think it might prematurely drive the price up.

      Bidding momentum is probably most likely to happen in that end-of-auction flurry of bidding since people might instantaneously revise what they thought the item was worth to them based on what others are bidding. It’s all in the heat of the moment so people might be more likely to get carried away in such circumstances. I’ve certainly been a bidder in plenty of auctions that have ended that way and I always have to remind myself that I need to stick to what I was willing to pay and what the item was worth to me, otherwise it really is too easy to get carried away!

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Post Navigation