A Few Thoughts on the Google Rating Guidelines

If you're in the SEO community, then there's one thing you'll learn quickly, and that is - What Google Says and What Google Actually Does are very often not the same thing. In fact, the worst kept secret is that while the Google public representatives and official blog posts will happily communicate with users on issues where it aligns nicely with the oft-touted "Don't Be Evil" philosophy, they can either be apparently misleading or, more often, completely silent on changes which don't align... quite so nicely, shall we say. It's also very common that comment is never passed on areas that SEOs know count as a major ranking factor, regardless of whether these areas will impact SEOs positively or negatively.

Such examples include backlink amount (traditionally important), linking across close IP blocks (don't do it to excess), backlink versatility (i.e. amount of uniquely linking domains - a much more important factor recently) and anchor text (if you've got too many "perfect" links as a part of your backlink profile, it's obvious to Google you've been heavily SEO'd, and they'll discount your links - leading to the charming acronym of BLOOP, or BackLink Over-Optimisation Penalty.

This means that if you want some deeper insight on algorithmic changes and how things really work over at the engines without doing extensive empirical testing of your own, the odd leaked piece of info is useful to either confirm or deny whether you're on the right track. Back in the day, AOL leaked some information on query strings and relative clickthrough for various positions, which I've seen used in various "traffic prediction" spreadsheets to this very day, even though it's now pushing 5 years out of date. This shows how valuable pieces of hard evidence are to SEOs.Which is why the Google Rating Guidelines for Search Quality Raters being leaked recently (and then subsequently removed from many sites in a frenzy of PR activity by Google) was a huge boon for many SEOs, as it confirms many things SEOs have already thought about the algorithm.

There are endless little technical details I could go into, but the big insight is that the constant refrain of "everything is algorithmic" is not quite true. Given the amount of raters inherent in the system, it's now fair to assume that if your site has any rankings for any major terms, then at some point someone is going to be casting an eye over it with the idea of assigning your site a good or a bad rating, and this does and will feed back to Google. Realistically, this was always going to be the case - with web design generally being such a chaotic discipline with a great many different ways to do the same thing (sorry Alex), there is no algorithmic substitute for a pair of eyeballs on a screen.So, what does this mean for SEO? Same as it ever meant - black and even grey hat just isn't going to cut it, because now more than ever you're going to get found out. Maybe not now, maybe not in six months, but when your site does get picked up for a query which looks a dodgy then you ARE going to get hammered in the rankings. Design your site for users, not the engines; don't cloak CSS, don't cloak javascript, and try to optimise your on-page stuff for user intent. Oh, and be really careful if you think your site sounds like it might have a vaguely racy keyword that could confuse user intent. Don't just put "breasts" as a deeper level keyphrase within a category if you actually sell "free-range organic chicken breasts", because "possible porn queries" appear to be a huge grey area.

Any comments or thoughts on the new guidelines, or our interpretation of them? We would welcome your comments below.