Shedding Light on Dark Social

How we're helping our clients see social traffic.


From Webtrekk Content Manager

David Vranicar

 

If someone wanted to play a joke on the publishing industry, it might look like this.

At a time when publishers are doing everything they can to make sure people share content, the act of sharing content is creating headaches for publishers.

Media outlets want you to share on Facebook and Twitter and Pinterest and Google+…


And on StumbleUpon and Reddit and LinkedIn…


And via email and WhatsApp…

Sharing is oxygen for today’s media industry. Sharing is also a problem for today's media industry. At least when it’s dark.

Why is it so dark in here?

When the term "dark social" was coined in 2012, it focused on email and instant message. The idea was that sharing over these channels was hard for analytics tools to digest, and was therefore logged as direct traffic (wrong) instead of referrer traffic (right).

No doubt, traffic that comes via email and IM is still tricky for analytics tools to sort out. But the problem is even more complicated today thanks to the proliferation of in-app browsers, which are every bit as dark.

If someone uses Twitter on a laptop, for example, then clicking on a link leads them away from Twitter and to a new tab in the browser. This generates metadata that is passed on to analytics systems.

Meanwhile, a link clicked within Twitter’s mobile app doesn’t send the user over to the publisher’s website via Chrome or Safari. Instead, a link within Twitter’s app opens up Twitter’s own browser.

As a result, no referrer info is sent from app to browser because the browser is the app. It’s a loop. A dark loop.

So analytics systems generally just scratch their heads – or do whatever analytics systems do when they’re confused – and mark this down as direct traffic, even though it is most certainly referrer traffic.

It would be nice to ignore this little analytics quirk – except that it’s not little. Various estimates put dark social’s share of social traffic at roughly 70%. In Europe, that number is 77%, almost five times more than Facebook’s 16%.

And this problem is only going to get bigger. Facebook and Twitter and other apps want to keep users within their own ecosystems, and in-app browsers are a great way to do that. Expect to see more and more in-app browsers from social media, from instant messaging apps, from email providers.

Let there be light

To circumvent this issue and give our clients a better understanding of their social traffic, we have configured our analytics software to detect the referrer app, even if that app has its own browser and therefore doesn’t send "traditional" referrer information.

It works by tracking the user agent string. Every in-app browser – Facebook, Twitter, all of them – identifies itself within a user agent string. This string contains data that we can use to “replace” the missing referrer information.

User agent strings are therefore the perfect tool – so far, the only tool – for shedding light on what used to be dark social.

Now, given how much traffic is generated by dark social, publishers’ reporting might look a little different after activating this sort of tracking. A marketing team that's really proud of all the direct traffic it is generating might be startled when the numbers crater.

But remember that the total traffic is still the same. It’s just looks different with the light flipped on.

 
GET IN TOUCH
Let's talk