Any marketer worth his or her salt spends a good amount of time looking at their web analytics and trying to figure out where their visitors are coming from, where they’re entering the site, how they’re navigating around once they get there and how long they’re sticking around.
Most of the people who do so are also likely frustrated every time they bump up against the “Other” category when looking at site referrals. Social networks are clear, as are direct links from other sites and other sources. But that “Other” section always frustrates since it can encompass so many things.
Alexis Madrigal at The Atlantic has confronted this issue head-on and dubbed it “dark social” since those sorts of referrals are hidden from view. He argues, convincingly, that almost 70% of social referrals are coming from sources that don’t fit with the current definition of “social.” Instead they’re coming from IM conversations, email threads and other communications that aren’t Facebook, Twitter or other social networking sites. Here’s what that looks like for The Atlantic in his story.
The main thesis of Madrigal’s piece, that the effort being put into optimizing content for sharing on Facebook and Twitter is only impacting a small percentage of web traffic, I think is a good one. And it’s interesting to think about this in light of the recent Buzzfeed study on what publishers have had the most engagement success on Facebook since when you look at the winners there they have the kind of content you “Like” or comment on right on Facebook as opposed to send in an email to a friend as something you really want them to read.
As Mathew Ingram at GigaOm points out in his own take on Madrigal’s article, the key to success comes back down to creating content that someone wants to not only read but also pass along. The mechanism they do so is almost irrelevant. It’s the responsibility of the publisher to create the impetus and provide the tools for someone to easily take action on that impulse. It’s then in the hands of the reader what tool they use to do so.