Jane Jacobs has passed away.

I’ve often sorted people into two groups. Those for whom the fact that the minimum width of a sidewalk is 12 feet seems obvious v.s. those who don’t get it. Jane Jacobs was in the first group. We will miss her.

Michael points out a nice article by Chandler Howell discussing ways that sellers can game eBay’s second chance mechnisms to assure they get the best price.

First off recall how an eBay auction works. The seller puts up his vintage pezz dispencer and the buyers put forward various amounts they are willing to pay. For example; $10, $15, $100. Ebay, in it’s role as market maker, computes the winning bid. In this case the guy who bid $100 gets it; but he only pays a small increment over the second highest bid.

That’s a nice example of the problem that frustrates all sellers. The buyers hide what they are willing to pay. If the seller in this example had known that the highest bid was $100 then he could have hired a shill to bid $99 and practically no risk.

In the liturature about pricing this is called the “hidden price” problem. I love that term because the first time encountered it I was reading about hotel room pricing and the paper went on an on about the problem of the hidden price and i thought the price in question was the minimum price the hotel would be willing to rent the room for; but no it meant the maximum price I would be willing to pay for the room.

EBay’s second chance program can be used to solve the hidden problem for the sellers. The nominal purpose of second chance is to allow sellers to sell their item when the buyer fails to follow thru on the purchase. For example he wins, but he never sends the money. This is probably reasonably common since buyer’s remorse is a natural part of the sales cycle. But second chance solves the hidden price problem by allowing the buyer to introduce a shill with no risk that the shill will have to actually buy the product.

The seller sets up his auction to assure he can do second chance – he pays a bit extra for this and does some extra work. For example he has to do a buy it now auction and then quickly have his shill make a low bid so the buy it now option is removed from sight. He then waits till the auction is about to end and has his shill make a very high bid. The shill wins the auction for slightly more than the value of the top real bidder. At this point the buyer knows the maximum bid of the high bidder. Latter he pretends the shill failed to pay and he offers it to the high bidder for that price.

Of course from the market maker’s point of view, i.e. eBay, the problem they are trying to address is that some percentage of auctions don’t work out. Buyer’s remorse is only one reason why a buyer might decide to fail to complete a transaction. Incompetence of either the buyer or the seller is a common reason. There is a marvalous term in the auction liturature called “winner’s curse” that points out that from a statistical point of view auctions are a bizzare way to discover the value of a item. Instead of getting the average of a large number of valuations the action coughs up the most outragous valuation. Buyers who win at an auction sometimes realize that and back out because they now know nobody on in the whole world of eBay things that pezz dispenser is worth what they just bid.

The seller like that curse. Using a shill and second chance can assure the winner is maximally cursed.

Spirit dancing suggests that my wife’s work triggers a reminding of Grant Wood‘s American Gothic. interesting.

I wonder if the big blogging sites have AD sales teams that place exactly the right add on blogs which are suddenly having their 15 minutes of fame. The volitility of long tail actors is very high; so it’s critical to catch those slashdot moments. Automated (look ma no hands) ad placment is all well and good, except when it’s stupid.

This is analagous to the puzzle of how to deal with the traffic peaks around slashdot moments so you don’t capture all the eyeballs you possibly can when they go by; but in this case the consequences go directly to the bottom line.

In both cases this is an example of why aggregating a stable of longtail actors can be critical to making the system function well.

While can imagine decentralized automation that quickly spreads traffic peeks across a collaborative group it’s much harder for me to imagine a scheme that brings AD placment talent to bear in a distributed manner. I can imagine that this is a feature that all the add networks would obviously have, or ought to.