Spam Wars - Pt 1

You should really only read this if you're very bored. It's meaningless and the real fireworks are going to happen in a few months - when new architectures for mail surface and are tested (and discarded) by the markets. Markets, plural, because the only real insight here is to remember that any viable mail architecture will be accepted by two kinds of consumers (or buyers and sellers if you want to model it that way): spammers and mail readers.

While I'm talking about spam, Eric Allman had an article titled 'The Economics of Spam' a few weeks ago, where he made some good points but got the econ backwards. The real economics of spam are about information - Bob Cringely made the real point here: that to stop spam, we have to make spam more effective, not less effective. Spammers don't spam because their marginal cost is zero (it isn't - addresses to spam cost very real money), but because their marginal benefit is incredibly low: they never know when they'll finally make a hit, and on average, they almost never do. Put another way, the point is that the rate of effectiveness for spammers has to go way up for spam to effectively stop.

If you don't follow this, think about it this way: what's ended the Red Queen race in web advertising? Models like Google's, which increase effectiveness rates, making it unnecessary to enter the arms race of flying logos and massive interstitials.