The way in which search engines determine how to rank documents is something that the search industry has been debating over for ages, but we may be about to be given a bigger clue than we’d ever of imagined.
Last week, I was listening to a web and software developer talk about how a maintenance script he had built will identify URLs within a CMS and redirect versions that it sees as duplicate as long as it has no conflicting information that would give reasonable doubt and require the script to make a judgement.
And that really got me thinking…
I recently wrote an article for Econsultancy about the Knowledge Graph and during my research I had become enthralled with the whole idea of a semantic web, particularly in relation to search. The idea that a computer will be able to call upon the vast data on the web and be able to make complicated analyses to create an answer to a query is as exciting as it is scary (anyone who is a fan of the Matrix films will share in my fear of the “machines”).
In their most basic description, search engines currently rely on their algorithms to determine relevance of a page and compare that to a number of metrics which identify how authoritative that document is and rank them accordingly. Now with a more semantic mindset Google has the potential to take this step further.
Just as the CMS script, Google will find URLs, yet it relies on inbound links and other signals (which we all know can be manipulated) to determine what sort of weighting to give to the information within. Like the CMS script, Google doesn’t want conflicting information clouding it’s judgement.
So what if it could use something like AuthorRank to establish the creator of the information, and weigh their personal profile including experience, online activity, relevance and numerous other data points to determine a more comprehensive analysis of both relevance and authority.
Sounds good eh? Less room for manipulation. Higher quality results, and another avenue of “understanding” for machines via the basic taxonomies needed to set up this sort of system. Yet as it stands there is no such thing. As is stands Author Rank is just a myth. Or so I thought…
A few days ago I was browsing Google+ (yeah people actually do that) and I came across a post by patent loving SEO (and a really top helpful guy) Bill Slawski about Google+ Sign-in. After reading a little more into the new authentication toy Google has released to developers, it became clear how important this is to not just their social network and app community, but also their search product.
Google+ Sign-in gives the search giant more of an ability to give everything you do online a social layer. It will be there when you’re writing your blog post about your brothers upcoming birthday via rel=author, it will be there when you leave a comment on your favourite website or buying a birthday present from your favourite store via Google+ Sign-in, it’s there when you upload your party photos and share a status to your friends to Google+ (or potentially other social networks if they adopt Google+ Sign-in too).
Google+ Sign-in is one of the missing links we’ve been looking for in the Author Rank conundrum as we can now see how Google will begin to evaluate users activities online to produce an estimated “worth” to their writing.
Still think Author Rank is a fad? I don’t.
Have some thoughts on the issue? You should leave me a comment below.