Googlenomics, Auction Theory, and Hal Varian

Like many programs, the Unknown Alma Mater's doctoral program used Varian's Economics text for our Micro Sequence. So, I perk up when I hear Varian's name mentioned.

It turns out he's Google's Chief Economist (who knew - Google has a chief economist?). A recent Steven Levy piece in Wired magazine talks about the ways the company uses economic theory (and auction theory in specific) in their Google AdWords program. Here are a couple of snippets:

At the time, most online companies were still selling advertising the way it was done in the days of Mad Men. But Varian saw immediately that Google's ad business was less like buying traditional spots and more like computer dating. "The theory was Google as yenta—matchmaker," he says. He also realized there was another old idea underlying the new approach: A 1983 paper by Harvard economist Herman Leonard described using marketplace mechanisms to assign job candidates to slots in a corporation, or students to dorm rooms. It was called a two-sided matching market. "The mathematical structure of the Google auction," Varian says, "is the same as those two-sided matching markets."

Varian tried to understand the process better by applying game theory. "I think I was the first person to do that," he says. After just a few weeks at Google, he went back to Schmidt. "It's amazing!" Varian said. "You've managed to design an auction perfectly."

...

AdWords was such a hit that Google went auction-crazy. The company used auctions to place ads on other Web sites (that program was dubbed AdSense). "But the really gutsy move," Varian says, "was using it in the IPO." In 2004, Google used a variation of a Dutch auction for its IPO; Brin and Page loved that the process leveled the playing field between small investors and powerful brokerage houses. And in 2008, the company couldn't resist participating in the FCC's auction to reallocate portions of the radio spectrum.

Google even uses auctions for internal operations, like allocating servers among its various business units. Since moving a product's storage and computation to a new data center is disruptive, engineers often put it off. "I suggested we run an auction similar to what the airlines do when they oversell a flight. They keep offering bigger vouchers until enough customers give up their seats," Varian says. "In our case, we offer more machines in exchange for moving to new servers. One group might do it for 50 new ones, another for 100, and another won't move unless we give them 300. So we give them to the lowest bidder—they get their extra capacity, and we get computation shifted to the new data center."

Read the whole thing here.

Winner's (and Loser's) Curse with Swoopo.com

Winner's curse is the well-known phenomenon where the winner of an auction is often just the person who's most likely to have overpaid for the item in question. So, often the winner is the loser.

Then I heard about Swoopo.com. It's been called "pure distilled evil in a business plan". Here's their setup:

• Bidding for an item starts at $0.15
• Each bid raises the price by $0.15
• Bids cost $0.75 to make.
• Here's the kicker - a bid in the final seconds extends the auction for 15 seconds. So, auctions can go on and on.
  

Of course, they post the "savings" you would receive if you bought the item at the current price as a prod for people to continue betting.

They also hold "penny auctions" on their front page - a bid only increments the price by a penny. I recently saw a TomTom GPS sold for about $12. That means 1200 bids at $0.75 per bid, for revenue of $900, on something that costs them between $300 and $500. Not too shabby.

This is a behavioral economist's dream - it has bidders focusing on sunk costs (I have to make back my bids, and if I win, I get the savings), hubris, and endowment effects (the bidders start viewing the item as "theirs", and therefore value it more highly). And if I thought about it a bit, I could probably come up with other behavioral biases.

It has some similarities to a "dollar auction". In this setting, individuals bid on a dollar. But the catch is that the second highest bidder must also pay their bid, but without getting the dollar in exchange. So, the second place bidder continues to escalate to cut their losses. In Swoopo's setup, the 2nd place bidder isn't obligate to pay, but once they're in the game, they continue bidding to recoup their already-paid bids (that is, if they can get the item at a discount).

It's not a scam per se, because everything is disclosed up front - the rules are clearly stated. But the only advice I can give you about using Swoopo comes from War Games (the 1983 movie starring a very young Matthew Broderick):



UPDATE: Here's a perfect example of how the irrational bidding that can take place - a person "won" an auction for a Sharp 42 inch LCD TV. According to the website, it was "worth up to" $1,199. The "winning" bidder paid $3,360. Assuming that this was a "normal" auction with bid increments of $0.15, this means that Swoopo received total bids of (3,360/0.15) x $0.75 = $16,800 in total bids (including $1,512 from the "winner" alone), in addition to the winning bid of $3,360, for a total of $20,160 -- all for an item worth at most $1200.

UPDATE2: as pointed out by a reader, the auction above was a "fixed price" one where the "winner" got to purchase the item for $119. So, the buyer could have potentially gotten it for a very nice proice. However, they ended up spending over $1500 in bids, plus the $119 winning price, all for a TV that was worth at most $1200. So, the winner endend up overpaying by at least $400, and Swoopo made total revenue of almost $17,000 for the cost of a $1200 TV.

One part of me wishes I'd thought of this - in about a couple of days time I'd have made back all the money my retirement accounts lost this past year. But I'd feel a little bad getting that money from stupid people. Not to say I wouldn;t do it, but I'd feel a little bad.