Sunday, January 4, 2009

SEO Science


Search Engine Optimization is still in the early stages of turning hypotheses and theories into a form of science. SEO science in my opinion delve beyond just purely stating what has happened or taking guesses, but rather using complicated econometric formulas for regression analysis via panel data. Of course, the idea of running econometric analysis appeals to a very niche audience as correlations and marketing proofs are usually preferred (and in most cases is all that is needed).

Nonetheless, there are consistent thoughts and ideas inside the SEO world on what works for proper optimization and on what search engine itself. What interests me most is algorithmically figuring out what kind of factor affects search engines the most when you can hold other values constant and by what percent. It is within the analytical area of SEO that interests me and not just the marketing and strategy approach (which obviously are just as important).

Effective analysis of the science of SEO would have to take in a large amount of factors that could influence Google, not to mention the areas where Google manually changes its algorithm (a large error factor). Various blackhat and grayhat techniques in heavy traffic and niche markets could easily create a large statistical error, but even more importantly, the constant changes in search engine algorithms would make econometric analysis a limited benefit if not re-analyzed every algorithm update.

Fear not, however, as this has never prevented people from doing time-series analysis for the stock markets and successfully predicting, albeit briefly, where to put one’s money.

1 comment:

john peter said...

manual directory submission service This is a really good site post, im delighted I came across it.