Google Indyrank Theories 15 July, 2006 — Stuart Brown

More to page ranking than meets the eye?

Posted in Search Engine Optimisation, Google
Tagged with: , , , , , ,

Google watchers everywhere were given a glance under the hood, as a blogger revealed a rarely-seen error message. Most of the output was as you'd expect, with debug status information and various parameters - but one thing mentioned was the mysterious 'Indyrank'. What could this new and previously unmentioned ranking variable refer to?

Semantics

Of course, we're dealing in nothing but speculation here - but the error message in question has been confirmed by Google employees to be genuine, but they're keeping tight-lipped over the exact meaning of them - perhaps the error message reveals more about the inner workings of the ranking engine than they would prefer us to know?

Of all the interesting parameters and terms, 'indyrank' captured my interest the most - and not just because I'm a fan of the Indiana Jones movies, either. So what could the term mean?

Personally, I suspect it's rather more likely to mean an 'independent ranking', with Indyrank simply being the snappier term for it. The question is, of course, independent from what?

Independence Day

Traditionally, the Google rankings have been determined by a complex algorithm based on off-page linking and on-page keywording and structure. This has been established since Google first launched - since the days of 'Backrub' and the idealised pagerank model.

Of course, any system that relies on purely technical factors can easily be fooled - and indeed, there's a whole SEO industry intent on foiling Google's ranking mechanism to achieve the highest rankings with the least amount of work. Spammers and affiliate sites will always find a way to break through filters designed to kerb their proliferation - their livelihoods depend on it. But these spammy sites bring down the quality and relevance of Google's sites, and so Google have presumably found a way to reduce the ranking of those pages in other ways.

The best possible judge of a website's 'quality' or relevance to a search is a user - a real flesh and blood human, with a brain and suchlike. Google must surely know this, and must by now employ some method of collecting metrics regarding real human usage of sites.

We know Google are collecting click information - so they know which results garner more clicks and interest from the user, and coupled with the Google Toolbar, they can measure if that person 'sticks' on that site or comes back to the search page - which could indicate that the page didn't meet the user's query as well as it could have done. There are more human metrics they could be taking, as well - with services such as Google Analytics installed on a large number of websites, they can get information on the best keywords for that particular site, and adjust the rankings for each accordingly.

Add to this the Google Secret Labs, which evidently uses real humans to train search quality filters, and you have the components of what could be a very important part of Google - a human-derived page ranking scoring, independent of the concrete world of linking and keywords.

Humans not Robots

Is the Indyrank score based on real-life human metrics? Is it a major factor in the rankings? As ever, it's difficult to say, but one thing is true - Google want to provide its users with the most relevant content, and what better metric is there other than the users themselves?

Google are quite clearly amassing data on user-pattern behaviour - and if they aren't feeding this data back into the system, then it's going to waste. This 'Indyrank' is a lot more difficult to fake - although presumably not impossible, which may be why they're keeping the exact details secret. This drive towards human metrics may also explain the incentives in place for Google Toolbar referrals - why pay $1 per Google Toolbar installation unless you plan to get something out of it?

One thing is clear, at any rate - in the future, SEO will be less about links and keywords and more about actually providing humans with a site with relevant information and 'stickiness'. It's a point that white-hat SEOs have been pushing for a while, and it seems Google may be making headway towards rewarding this behaviour.