If you want to be mesmerized by a fascinating auction site, go to Swoopo.com and click on an item that has less than 15 seconds on its countdown clock. Swoopo is what some of my colleagues have recently termed a pay-to-bid auction. You pay $0.75 to bid, and each bid raises the price by $0.15. If the item has been up for auction for more than 24 hours, then a 15-second timer gets reset every time somebody makes a bid. If no one else bids within the 15-second period since your bid, you win the item.
Two of my colleagues, Brennan Platt and Joe Price, along with one of our undergraduate students (Henry Tappen), have written a paper entitled, "Pay-to-bid Auctions," that beautifully lays out the economics of these Swoopo-type auctions.
Two of my colleagues, Brennan Platt and Joe Price, along with one of our undergraduate students (Henry Tappen), have written a paper entitled, "Pay-to-bid Auctions," that beautifully lays out the economics of these Swoopo-type auctions.
Before getting to the results of the paper, one of the greatest strengths of this work is that it combines solid economic theory with great data work. Henry Tappen wrote code to collect the data on thousands of Swoopo auctions directly from their website in real-time. This provided a fabulous dataset with which to test their model.
Aside from the clever data collection, this paper has at least two interesting results. First, they are able to match the distribution of ending prices (which directly reveals how many people bid on the auction and total Swoopo revenue) for the majority of items Swoopo auctions with a simple model that uses the market value of the object. Interestingly, the data and the model show that Swoopo does not make large profits on most of its auctions.
The second interesting result is that their simple model cannot match the distribution of ending prices for a specific class of items--video game related products. In particular, the distribution of ending prices on video game products has much more weight in higher prices than the model would predict. Platt, Price, and Tappen find that if they allow for slightly risk-loving preferences, they are able to match the video game items end-price distributions as well. They also provide some independent evidence that gamers do in fact have risk-loving preferences. If this is the case, then the Swoopo business model might be to use the more conventional items as loss (or break even) leaders in order to mask the excess profits from its auctions of video game items.
In the end, Swoopo's pay-to-bid auctions present fascinating incentives and dynamics that are laid out beautifully in Platt, Price, and Tappen "Pay-to-bid Auctions."
Aside from the clever data collection, this paper has at least two interesting results. First, they are able to match the distribution of ending prices (which directly reveals how many people bid on the auction and total Swoopo revenue) for the majority of items Swoopo auctions with a simple model that uses the market value of the object. Interestingly, the data and the model show that Swoopo does not make large profits on most of its auctions.
The second interesting result is that their simple model cannot match the distribution of ending prices for a specific class of items--video game related products. In particular, the distribution of ending prices on video game products has much more weight in higher prices than the model would predict. Platt, Price, and Tappen find that if they allow for slightly risk-loving preferences, they are able to match the video game items end-price distributions as well. They also provide some independent evidence that gamers do in fact have risk-loving preferences. If this is the case, then the Swoopo business model might be to use the more conventional items as loss (or break even) leaders in order to mask the excess profits from its auctions of video game items.
In the end, Swoopo's pay-to-bid auctions present fascinating incentives and dynamics that are laid out beautifully in Platt, Price, and Tappen "Pay-to-bid Auctions."
Fascinating. I'm not a mathematician or economist, but I enjoyed reading both your blog post and the paper. As far as I can tell, it looks like it is logical to participate in the Swoopo scheme, as long as you do it right. It's just figuring out how to do that which is the trick.
I agree with Brendan on this. If you do it correctly, you will end up winning an item at a very low price. You also have to rely on luck though.
I think the business model is ingenious. However, I would not participate for a good number of reasons.
Being that the Business itself is making money off the bids, I would feel there would be less likelihood for any internal monitoring system from preventing a self-outbidding scheme from taking place.
Lets look at it this way.
Swoopo puts up a hot item... lets say a 50" Plasma TV. Bidding opens up at $1.00 with a half hour bidding time.
It gets a few bids and by the time it reaches a 5-Minute marker it is still priced rather low at lets say $27.00.
Now is where the "Fun" begins. With time breaking 5-Minute marker, the bidders watching the auctions are getting more tempted to closely watch the item to try to "pick" their moment, hoping no one else is going to bid.
By the time it breaks 1 Minute remainder time on auction, the 50" Plasma is up to $50.00. If it was a $0.01 auction item, each bid was increasing the auction price by $0.01 / increasing seconds on item by 1-4 seconds (from what I witness watching an auction today). Each bid transaction comes at a cost of $0.60. For the item to have reached a price of $50.00, there would have been4,900 bids @ $0.60 a bid for a total item sale value of $2,940.00 for that 50" Plasma. ($49.00 / $0.01 x $0.60 = $2,940.00) I doubt they paid retail value for that 50" Plasma, so the percentage of Markup on the item sold even at $50.00 would be well over 100%.
That's just the icing on the cake.
Now lets say Joe Blow was banking on getting himself a cheap Plasma and this looked like too sweet a deal to pass up. Joes been bidding since the minute marker and made a grand total of 30 bids (30 bids @ $0.60 = $18.00). Lets be honest, $18.00 isn't that much money really, but if Joe doesn't win the bid, he spent $18.00 on absolutely nothing. Compound that by the thousands and you really got a huge rip off scheme going.
And I am hardly finished. What Joe Blow doesn't realize, along with the hundreds of other participating bidders, is that Swoopo stacked the decks! They've got auto-auction bidders who will continually outbid you on every item till the final tick of the clock. “10 seconds, 8 seconds, 5 second! What, Outbid by JohnnySpot2 yet again!!!... I’ll win this TV YET!”
As I said earlier, who is monitoring these auctions to ensure that this is not happening and how is someone to know they auctions are legit? Who is to say that Swoopo hasn't set into place an auto-bidding program behind their website for every available auction, or that they aren't hiring people on the side to bid on the items themselves, and paying them a percentage of the bottom line $$$ markups they are making off of items. Who is tracking the items sold to ensure there are actually items being delivered to "customers" as you might think of them.
I highly doubt there is any monitoring of a site in this nature, and I am willing to bet that there is a indecipherable user agreement the length of a full series of encyclopedias covering Swoopo's A$$ for the legality of the business they are conducting.
If that penny auction 50" Plasma went up to $200.00 on the final sale, Swoopo brought in an astronomical $11,940.00 for a set that would have retailed $1000.00 - $2000.00 in stores today.
If Swoopo is engaging in self bidding on auctions, and they do not actually let a real bidder/customer win an auction, that $11,940.00 would be ALL PROFIT, since in fact no item was sold or delivered to a "customer"/bidder. I don't care how they may be doing it, but I would be very skeptical on the ethics of this business and would lean towards the angle that this may in fact be the very thing they are doing.
Someone ought to have the business investigated for fraud.
Good deal or not for an individual purchase, this is capitalistic mentality at it's dirtiest.
Hey... SWOOPO... CUT ME IN!!! I'll bid on your auctions for you all day for a measly 1% of your profits from the auctions I partake in, and laugh my A$$ off all the way to the bank just like you are doing.
Don't forget the "Swoop it Now" feature that let's people put the money they spent bidding on an item they didn't win towards buying that item at full retail from Swoopo. So not only does Swoopo make money off the original auction, but also lots more from the people who lost out, feel like they wasted money and turn around and buy the item for several times what they thought they'd pay. Crrrazy!
Sites like Swoopo are more for fun than for people actually looking for a bargain. I suggest you look around at the other auction sites similar to Swoopo. Hasteno.com is a new one that I’ve been playing on and I’ve actually won a couple things. Playstation 3 games going for $.08 (2 bids!) and a Nintendo Wii going for $2.16 is ridiculous. Before you completely discount the idea, you should check them out.
Hey all!
I'm knew to all of this and have posted on a couple other sites regarding some information so figured I'd try again here in hopes of getting a response.
This past week I found this new site, www.BIDZONE.com, and it looks pretty interesting. The product price actually decreses everytime someone places a bid on the item??
Has anyone else heard of BIDZONE? Let me know!
I can't believe that the nintendo Wii and DS have way more sales than the PS3 and PSP.