Typo has fairly impressive SEO

November 14, 2007 at 9:43 AM

I won't pretend I know much about SEO. I've read the original PageRank paper, and several others offering criticism or improvements. I've even implemented a rudimentary Ruby library to calculate such rankings, but PageRank (as far as the mathematics go) only describes a system for ranking based on the voting relationship of a directed link graph.

SEO, on the other hand, involves divining the black magic of how a given engine (usually Google) interprets the text on a page. Certain terms are given different weights depending on frequency of occurrence, location of occurrence (inside which tags? et cetera), and other unknown factors. Simply put, the pure mathematical PageRank model doesn't translate to the reality of Google's search implementation; at some point, the PageRank and the term weights are mashed together in an inverted index, using a recipe that only Google knows. And frankly, I'm not that interested in figuring out exactly what that recipie is, or how to use my knowledge thereof to make my little blog rise the ranks.

Despite my indifference, I was pleasantly surprised to have a brief (1 day) reign in the #2 spot for the query unreleased u2. I was dethroned after a mere day, although (as of 2007-11-14) I'm still on page 2 of a different query. Of course, the real goal is to have iLike's page with the actual video content of Bono singing an unreleased U2 song have a prominent position in search results. I attribute my brief success to the great way Typo organizes posts. Way to go, guys!