SEO: Art or science?

Disclaimer: I, or Connected, do not sell or market any SEO services. Ever.

We’ve been around a while, in fact we’ve been around pretty much since the explosion of the Internet in the late 90’s and during that time we’ve seen our fair share of snake oil come and go. There is one subject that stirs up more than it’s fair share of opinions and that’s SEO…or Search Engine Optimization if you’ve lived under rock for the last 10 years.

Never been a fan really, pretty much decided early on that if you follow the Search Engines’ (read: Google) guidelines and build sites using best practice then you’ll work out nice and dandy. But there does seem to be a huge following and the most diverse advice in this field, so much so that almost every SEO ‘specialist’ will give you different and conflicting advice (for a price, mind).

I do accept that there are some tricks and neat stunts that, to a certain extent, can pull the wool over the engines’ eyes but not usually for long and not usually with any kind of long-term gain. So, making me chuckle this week, then, is SEOMoz telling us all that there is a science to how well you rank. Really? Fantastic!

Except, if you elect to use science to ‘prove’ something then you had better get the science right. SEOMoz may have misunderstood statistics and, worse still, cloaked this bad science in sciencey-sounding words like Mean Spearman Correlation. This has generated a flurry of arguments discussing the various merits of using different methodologies to analyses data. Worth the read, but only if you’re secretly hiding a love of stats.

Some nice rebuts too from Ted Dziuba and Dr E. Garcia – both who warn about paying attention to this kind of science. Alternatively, you can read how Google suggest you go about working with search engines (PDF).