Colleges are committed to social media. Students report that campuses are using social media to help them connect with each other and student organizations, to provide information about education programs and career opportunities, connect with faculty, and even get information about financial aid. Many campuses view social media as an important tool to drive website traffic. Yet, according to a recent benchmarking report from Higher Ed Experts, only 4% of traffic to college and university websites is referred by social media. Or is it?

I believe colleges are underestimating the impact of social media on website visits.

Why? Dark social and the continued rise of mobile browsing.

If you’re not familiar with the concept of dark social, here is an excellent primer by Alexis Madrigal, former senior editor at The Atlantic. Basically, it’s the idea that a lot of social sharing happens outside of formal social media platforms. We email links to our friends and colleagues, or send it in an instant message. Web analytics tools do not recognize these as social media referrals. They most often recognize them as direct traffic.

The Higher Ed Experts benchmarking report shows a whopping 35% of all college & university website referrals are direct traffic. Are we really supposed to think that 35% of campus website visitors are typing a URL into their browser or using a bookmark? I think that’s highly unlikely.

The plot thickens when you think about the rise of mobile browsing, particularly on Facebook. In 2014, Facebook reported an average of 890 million daily active users—and 745 million of them were mobile users.

That’s 84% of daily active users choosing to access Facebook from a mobile device, likely through the Facebook app.

So, what does this have to do with dark social?

Mr. Madrigal published another excellent article in December 2014 that cracked this idea wide open. He conducted an experiment that showed that users that click on link while using the Facebook mobile app are tagged as direct traffic in web analytics software. This is traffic we’re supposed to assume comes from a visitor directly typing a URL into a browser, or using a bookmark… or visiting from any source that isn’t correctly tagged in the analytics platform.

After a lot of nerd business, Chartbeat, a web analytics provider that competes with Google Analytics, figured out a way to correctly attribute Facebook mobile app referrals by parsing the user agent passed along with other visit information. As of December, many websites are now using this technique to more accurately report referral traffic from Facebook.

The Atlantic used this measuring technique and saw direct traffic drop by 40%, and Facebook referrals increased by a corresponding amount. Obviously, college & university websites are not exactly like a major media website, but most are news-producing properties that expect to drive traffic from social referrals.

The Higher Ed Experts benchmarking report is based on Google Analytics data, and currently Google Analytics does not use the same measuring technique for Facebook mobile referrals that Chartbeat does. So, if we were assume the same change in direct traffic and Facebook referrals for college and university websites, direct traffic would drop to 21% and social media referrals would jump to 18%, with the majority of that traffic coming from Facebook.

If you were able to show that Facebook drove more traffic to your website than all other referring websites combined…

…Would administration take your social media efforts more seriously?

I think so. I ran this idea by Joshua Dodson, the smartest web analytics guy I know in higher education, and he mentioned that without fancy coding, you can use the Full Referrer Report in Google Analytics to see if your site is receiving Facebook referral traffic that is not appearing on the social media overview or Referral Traffic report. When he compared the Full Referrer Report to the Referral Traffic report on one site he manages, he saw a 5% increase in reported Facebook referrals.

If you have access to a web analytics tool that correctly categorizes Facebook mobile referral traffic, or have a chance to run the Full Referrer Report in Google Analytics, I’d love to know what your social media referral numbers look like!


Article Author

Liz Gross

Liz Gross is a Social Media and Market Research Strategist for a federal student loan servicer and teaches Social Media Measurement for Higher Ed as a member of the Higher Ed Experts faculty. She’s also the author of How to Manage Social Media in Higher Education. Liz is a candidate for a Ph.D. in Leadership for the Advancement of Learning and Service in Higher Education at Cardinal Stritch University. Her dissertation research examines the impact of communication method on the frequency and content of college student interactions with faculty. She blogs at , and you can find her on Twitter at @LizGross144.

  • Erik Hagen

    Very interesting….I ran the Full Referrer report and I see a 67% increase in reported Facebook referrals over the past month. I searched for “facebook” within the results of each report to find all the variations. In the standard Referral report I get 4 domains:,, and In the Full Referrer report I see all of these as well as just “facebook” which has a substantial amount assigned to it. I wonder where that’s coming from…perhaps the mobile app user agent?

    • Liz Gross

      I don’t see that same pattern on my Google Analytics accounts, but they’re just for my personal sites, not a large institution. Can you add the secondary dimension of “device category” to see if all of those “facebook” visits are from mobile devices?

      • Erik Hagen

        So yes, they’re pretty much all from mobile, and applying a Campaign dimension shows they are tied to Facebook ads being targeted to mobile users. The source/medium is “facebook/social”. So I can throw those out since they aren’t real referrals.