There was a bit of an (ahem) stir earlier this week in the blogosphere over a post by respected ZDnet blogger Donna Bogatin, where she wrote about some comments on click fraud made by Google CEO Eric Schmidt.
While she reported Eric’s comments accurately, her blog lead off with the startling headline “Google CEO on click fraud: 'let it happen' is perfect economic solution”. Her blog got all kinds of furious reaction in the blogosphere, with people blowing up about how terrible Google is, etc. Google ultimately responded here (via Battelle).
I have examined the actual talk he gave and have come to the conclusion that he was misunderstood.
First, understand the context of his remarks: Schmidt, himself highly educated, was giving a talk at SEIPR, the Stanford Institute for Economic Policy Research, a group of smart people who discuss and work on matters involving theories of markets and the like. It’s not a talk with a group of Wall Street analysts, who want facts and figures.
At the forum, he was asked a question by a Stanford student which came in two parts:
a) What are you thoughts on click fraud?
b) Is there an economic solution more than just technological solution?
Schmidt answered the first part of the question by saying that:
1. Google worries about click fraud a lot
2. They have engineers who think it’s “great fun” to get ahead of click fraud (in this case, he didn’t mean “great fun” as marginalizing the issue — it’s a phrase more in keeping with the spirit in scientific and academic communities to look at deep problems with a sense of game).
3. Click fraud is not that large of an issue for Google.
4. Google does, in fact, try to detect it and eliminate it, and tries to decrease the time and increase the rate by which they automatically detect it — a process Schmidt refers to as “a good economic answer”.
Then in answering the second part of the question, Eric waxed theoretical on something like a form of efficient markets, where he said, “let’s imagine, for purposes of argument” that nothing was done about click fraud, there would ultimately be “a self-correcting process” (basically, advertisers will pay less for ads because the quality of ads are being diluted by click fraud).
While econ wonks can argue about his point ad infinitum, citing asymmetrical information theory and the like, one can see that his point is, in fact, fairly reasonable — on a theoretical basis, which is exactly how he framed the answer. As a crude example, look at the market for post-Katrina housing right now in New Orleans. The price of the houses already reflects the fact that they are a) mostly destroyed and b) have a risk associated with future hurricanes. Yes, I know there are other sides to this perhaps overly-simplistic view but it’s a long and exhaustive discussion perhaps relegated to another arena.
In a world increasingly demanding hard facts and answers in some kind of obsessive Teutonic fashion, a world which operates on sound bites, Schmidt made the mistake of exercising his right to talk in abstract terms. Now, the content of his economic theory may be under debate, but in no event did he actually state this theory as Google policy. And for that the uproar? Is it not just possible to simply speculate or theorize publicly in the subjunctive, without worry that words will be taken out of context and transformed into the indicative?
You can see Schmidt’s actual words here — just forward to 31 minutes (or listen to the whole talk, it’s quite interesting).