Statistical SEO

One of the things still missing today within the SEO community is a strong understanding of quantitative analysis and econometrics along with the actual desire to publish a paper within a constantly changing field. SEO is heavily focused on marketing (as it should be), but the idea of figuring out what works, by how much, and for how long seems to be sorely lacking.

In my view, statistical SEO is an open area that can be filled through a kind of econometric SEO, to help reduce the pervasiveness of common myths and rumors that often seem to spring up out of sites such as WebMasterWorld where anything that can go wrong is some kind of penalty rather than actually quantitatively testing what may have occurred.

You will see within this blog various ideas and concepts that will be driven not just out of a marketing perspective but also from a statistically-minded person probably in the forms of equations like the following:

OrgTraf_it = α + β₁*SERP_it + β₂*SEO_it + β₃*(SEO*PPC)_it + … + uit

And coming from the school I learned econometrics from, as far as I am concerned, if the econometrics formula for SEO is not using panel data, then there is no purpose in running the statistical regression for causal purposes. So please remember, correlations do not equal causation!

Therefore, when trying to figure out the right ways to get more traffic and what areas to invest in the most time, only truly statistical SEO analysis can figure out that to a more hard science.