According to http://www.clickz.com/clickz/column/2318344/tracking-dark-social-media-a-light-at-the-end-of-the-tunnel you can use the following tools to track dark social sharing, Tynt (free software) or Rio SEO (paid for software) or you create an advanced segment in Google Analytics to further analyze traffic.
Extract below taken from http://blog.hubspot.com/marketing/what-is-dark-social-faq-ht and http://blog.hootsuite.com/everything-you-need-to-know-dark-social/
What is dark social?
Dark social is basically any traffic you get to your site because your analytics can't correctly identify the site sending you the traffic and accounts for up to 60% of overall referral traffic for various websites. . In simple terms, referral traffic is usually identified by certain “tags” attached to the link whenever it’s shared. There are two main reasons why your analytics wouldn't be able to figure out the exact source of a visit:
1) If the visitor is coming from a source that can't be traced by your analytics program.
This is what we were talking about: via text message while browsing on your mobile, one-to-one emails or your company's IM client or you viewing an article on another person's computer with them. Most often, the reason your analytics can't keep track of the source of the traffic is because the source doesn't send referral data -- which is how the rest of your analytics gets bucketed.
When someone clicks on your site's URL on Twitter, for example, Twitter will let your analytics know that that's where the visitor came from.
2) If someone shares a link with UTM parameters (those little strips of code you manually add to a URL to track traffic sources, among other things) on an incorrect platform.
For instance, if someone were to share a link through email with UTM parameters on it that designate visitors who click on that link as coming from Twitter, the person they shared it with would get bucketed into the Twitter source category ... even though they didn't actually come from Twitter.
What dark social is NOT is the "Not Provided" under your organic search settings. Though not knowing what keywords are sending you traffic is definitely frustrating, your analytics still knows that people are coming from searching Google.
Make sense? Now that we're on the same page about what the heck dark social is, I know what your next question is: Is there any way to shed light on your dark social to figure out exactly which sources are sending you traffic?
Really, there's only one way to decode dark social: by removing incorrect UTM parameters from URLs you share on dark social platforms. Yup, I know this isn't for your site, but think about how happy the marketer on the other end will feel when they realize their traffic is more accurate!
So go on, pay it forward. Removing incorrect UTM parameters by hand -- your fellow marketers will start to see their analytics become a little less dark. And who knows? Maybe yours will, too!
The origins of dark social
In 2012, Alexis C. Madrigal wrote an article for The Atlantic called “Dark Social: We Have The Whole History of the Web Wrong” In this piece, Madrigal tells the story of questioning the existing analytics about web traffic, and the fact that they don’t seem to account for sharing links over instant messaging, a communication tool that precedes social networks.
Madrigal turned to data experts at Chartbeat, a tool that provides real-time analytics to publishers, which led them to discovery of a mysterious source that accounted for nearly 60% for traffic on The Atlantic’s site—something Madrigal goes on to call “dark social,” similar to dark energy, a term that’s used to describe a hypothetical form of energy that permeates all space and accelerates the expansion of the universe. (Quite poetic if you apply the same idea to the social web, isn’t it?)
Aside from the fact that The Atlantic article is a highly interesting and relatively easy read, no matter your level of familiarity with different engagement metrics, it also makes two very important points about dark social. The first is the fact that the most important shareability factor in a piece of content is the content itself. No good content = no sharing, however sophisticated your optimization efforts may be. The second point Madrigal makes is that the emergence of social networks didn’t create the social web, but only structured the existing channels by the act of publishing—and tracking—our social interactions.
Seeing dark social in a new light
We’d be happy to leave it at that, but like many other things in the world of web analytics, the story of dark social is a lot more complicated. Just this week, Madrigal wrote a new article, where he revisited the metrics surrounding dark social referrers. His new theory around dark social involved a very much familiar player—Facebook, or more specifically, Facebook’s mobile app.
Since Madrigal’s discovery in 2012, Chartbeat has been trying to narrow down the sources of dark social. Chartbeat’s chief data scientist Josh Schwartz told Marketing Land that Chartbeat set up a test site to eliminate some possible dark social sources, and discovered that some major social networks are not consistent with attaching referral information. Links that come from Reddit on desktop and mobile browser, for example, are trackable; but Reddit links that come from top mobile Reddit apps don’t always contain referral data.
It also turned out that Facebook doesn’t always attach the referrer to the link when a user engages with it on Facebook’s mobile app, or when a desktop user opens the link in a new tab. Given Facebook’s permanent spot at the top of mobile apps, and an established position as a leader in news sharing, this source of “dark social” makes a lot of sense.
What dark social metrics mean for your engagement strategy
Before Chartbeat and Madrigal broke the news about the newly found origins of dark social, RadiumOne has released a helpful report outlining some trends seen in dark social sharing. For example, they found out that 32% of surveyed people only share using dark social: emails, text messages, IMs, or forums. Furthermore, some information tends to be shared privately more than others; top categories for dark social sharing are arts and entertainment, careers, travel, science and education. Plus, the study finds out that click back rates, or the number of times someone clicked on the link you shared, are very high on messaging shared via dark social means.
In light of recent discoveries, one of the things you can do to unmask the origin of dark social traffic is to check for a simultaneous spike in link traffic coming from Facebook or reddit. Major websites have also reported digging into user agent data, which includes a line of code users leave after visiting a website, which identifies their operating system and browser type. User agent information, while not always translated correctly by analytics software, can provide more details about the referrer.
Finally, as Madrigal pointed out two years earlier, “There’s no way to game email or people’s instant messages. There’s no power users you can contact. There’s no algorithms to understand.” The best way to create content that is guaranteed to be shared is by writing interesting, informative, original material.
