• by Erik Sass , Staff Writer, December 8, 2014

Recently a lot of attention has been paid to “dark social,” referring to visits to social platforms and sites connected to them, that can’t be accounted for with the usual Web measurement tools. There’s been all kinds of speculation as to its potential sources: is it people sharing stuff on email? Instant messaging? Private networks? Measurement error? Ghosts? Space aliens? (OK, maybe not those last ones).

Now one Web traffic tracker, Chartbeat, seems to have identified the source of at least some of the mysterious traffic: it’s from Facebook mobile app users, who aren’t tracked because the app-to-Web site handoff doesn’t identify a specific referring source. However according to Chartbeat and Alex Madrigal, the tech writer with The Atlantic who first coined the term “dark social,” these visitors can be identified with a fair degree of confidence from another piece of information in the form of the “user agent,” which helps with compatibility around browsers, operating systems and the like.

They began tracking the user agents and voila: mobile traffic from Facebook was a major source of dark social. Given these insights, in retrospect it’s no surprise the volume of dark social traffic has increased in step with the massive increase in smartphone ownership. Chartbeat and Madrigal Web assert that publishers and advertisers have probably been underestimating the true volume of traffic referred to them by Facebook via previously unidentified mobile referrals.

Chartbeat is setting out to shed some light on the subject of dark social with a new dashboard for its traffic measurement service that gives full weight to visitors coming from the Facebook mobile app. The mobile measurement should pin down the referring source of anywhere from 10%-50% of dark social traffic. The company’s chief data scientist, Josh Schwartz, tells Marketing Land that so far, sites are generally seeing the volume of dark traffic dip by 10%-20% when the new mobile measurement capabilities are activated, while mobile Facebook traffic jumped by around 40% on average.

As Madrigal points out, this also means that Facebook’s domination of referral traffic to online publishers is, would you believe it, even more total than was recognized before. It turns out that 800-pound gorilla was more like a 1,200-pound gorilla!