Digital Gaggle’s thought provoker of the day for me was a digital marketing rockstar, Majestic’s very own Marketing Director Dixon Jones’ talk "How Trusted Is Your Website? Find Out How To Score Every Page On The Internet".

Don’t you just love it when you start questioning something that previously felt like part of your mind’s furniture?  Dixon’s opening gambit about trust shone an incriminating blacklight onto a belief I had never questioned until I saw the stain...

Do I trust a search engine more than a person?

Because the answer is yes, I definitely do. 

This, particularly for me, is very uncharacteristic.  I don’t use social media, I block Analytics, I do my utmost to prevent Google from tracking my location, my search history, my habits, my relationships, my schedule.  As futile as it may be, I do my utmost to stem the inexorable torrent of inane - yet intensely personal – information that inevitably gets sucked into the datacentres at Mountain View.  Yet, I trust the search engine to answer my question more than I trust my acquaintances and friends.

Why?  One aspect is immediacy.  We are impulsive creatures, and the opportunity of being able to have something right now when we’re thinking about it is just too seductive to resist, compared with the alternative.  I’ll wait, and I’ll probably forget to ask.

Dixon rightly points out motive.  Now that I think about it, my friends (probably) have fairly good intentions, in comparison to the faceless, data hungry machinations of the world’s most sophisticated advertising engine.  What the hell is wrong with me?

What about Googling embarrassing questions?

Another part of it is shame.  Some questions are embarrassing to ask.  That doesn’t just mean questions about awkward bodily functions – some stuff is just embarrassing to admit you don’t know.  It can also be inappropriate to start questioning your friends on electrical engineering as they sob into their pint and mourn the tragic demise of their doomed, loveless relationship.  Tends not to go down too well…

Equally, just because they are your friends, doesn’t mean they aren’t idiots.  Friends can frequently be unreliable.  For years, I sincerely believed an anecdote told to me by a close friend, whereby an angry crowd booed 90’s one-hit-wonder pop group Toploader, demanding that they repeatedly play their track “Dancin’ in the Moonlight” repeatedly for the entirety of their set.  In my defence, the guy knew loads of obscure facts about music.  He was in the ‘trusted group’, so when it came to spurious but hilarious details, I believed it without question – until I heard myself parroting the story to someone else thinking “can this be true?”  I was also about 14, which probably didn’t help.

How many clicks to Hitler?

There is a sense that by submitting a query to Google, you are delving into a comprehensive repository of all human knowledge.  How can Jim from down the pub compare to that?

Challenging that assumption alone would have been plenty insightful, but the remainder of Dixon’s talk proceeded to explore the mechanisms by which Majestic (and, albeit using some other route but inevitably reaching similar conclusions: Google) determine these trusted groups.

By studying the link relationships of large sets of data (the larger the scale, the lower the error rate) they are able to reliably predict the trustworthiness of a particular page by its relational distance to other trusted areas of the index.

Even more usefully, it is this same mechanism that is used by Majestic to determine the topical trust flow and citation flow metrics.  Just as a page’s trustworthiness can be estimated (with a good degree of accuracy) by measuring the distance from the given page to trusted areas of the internet, so to can a page’s topic, and authority on the topic, be reliably estimated by looking at the distance between it and other trusted pages on that topic.  Is the page not related at all to other pages about cakes?  Then it probably isn’t about cakes!

Trust I seek and I find in you

Looking to the future then, Dixon predicts that we will increasingly see a web where trust and authority is derived from people and particular pages, over domains.  A link from Twitter is not inherently valuable, but a shout out from a particularly influential person on Twitter is.  As we move towards an ‘internet of things’ and Google becomes even better at tying together your website, your social profiles, your wikipedia page and tying them to a concept of you - and better yet, a concept of you as a person - the influential  among us will be presumably be able to wield that authority like a commerce hammer.

Except me obviously.  And if Dixon's (sometimes dystopian) vision of the internet of things is to be believed, I’d better invest in a torch, as I won’t even be able to turn on the lights in a few years.