The Spam Algorithm
In my email inbox today I got this typical spam message (Yahoo Mail, not Gmail, of course) :
SENDER: Cartwright G. Ranald SUBJECT: porch amorphous
I usually subconsciously mark these as spam and move on with my life, but for whatever reason, tonight I actually noticed it and laughed. So typical. So funny.
This spam formula has been around for quite some time… and despite the commonly held belief of the effectiveness of “genetic” / rapidly adaptive algorithms (and counter-algorithms), it seems that what works for tricking spam-protection systems doesn’t work for tricking people.
Similarly, in the world of search engine optimization (SEO), what works for tricking search engines often doesn’t work for tricking people. Just because a web page gets a high “natural” search result on Google doesn’t mean it got there “naturally” …and it’s obvious to a person looking at that page. Just as with spam, in SEO there is an algorithm vs counter-algorithm arms race going on.
Creators of spam and SEO’ed web pages are both playing a cat-and-mouse game with the guardians of our digital content — those who would protect us from unwanted information/noise. Why are the creators of SEO’ed web pages doing better than the creators of spam? There are some key differences between the activities of spamming versus SEOing. For one thing, the “evolutionary cycle” in the spam arms-race is instantaneous: a blast of email either makes it through the software filters and gets delivered, or not. In comparison, the SEO arms-race is a geological phenomenon: it can take three to six months to see the result of one’s efforts. Also, spam is entirely mechanically mass produced while SEO is largely a hand-crafted (or at least highly customized) blend of art and science, with a few mechanical tricks here and there. You’d think that spam practices would have evolved faster than SEO techniques. So does the apparent success of SEO algorithms vs spam algorithms mean that people are, in fact, smarter than robots?
It seems to me that the two-way evolutionary pressures in the SEO world are a little more complex and nuanced than those in the one-way broadcast world of unsolicited email. It’s easy to know whether an email is “relevent” because if you didn’t ask for it and you don’t know the sender, then it’s probably not. In contrast, it’s hard to know the relevence of a search engine result given the open-ended nature of keyword-driven queries and the weakness of the search algorthims themselves. In a sense, in the SEO world, the reader is a participant in the SEO process. Between the reader, the SEO’ed content provider, other sites/blogs/link-clouds and the search engine itself, there is a combined (but not coordinated) effort to bring the most relevent web pages to the top of the list.
So the reader can’t always tell if (or how) a web result has been SEO’ed, versus how obvious it is in the case of email spam.
Meanwhile, a bigger question is why hasn’t Yahoo! figured out how to screen these emails from Cartwright G. Ranald? It’s not like Google is staffed by rocket scientists…