Rebooting Twitter's analytics

There's no denying Twitter's massive reach into the b2b marketing world; but, until recently, it was hard to tell just how far that reach really went if you used Google's analytics to track your Web traffic.

If you noticed an increase in your Twitter traffic over the past few months, there's a good reason why—and unfortunately, it might have nothing to do with your useful tweets.

The issue has to do with the way Twitter referrals were tracked through popular third-party applications such as TweetDeck and HootSuite. When a user clicked on a link embedded in one of these applications, the referral to your website wasn't recorded as a Twitter referral. Instead, Google recognized it as a “direct referral,” thus hiding some of the Twitter-related traffic under the direct referral column on your analytics report.

“The traffic is the traffic regardless of the source, but it was underreporting the source,” said Chad Pollitt, director of inbound marketing for Kuno Creative. “Because TweetDeck is a software application that sits on your desktop, when someone clicked on a link on TweetDeck, it was reported as a direct link.”

Last May, however, Twitter purchased TweetDeck and set about fixing the issue. Twitter's fix involved the gradual introduction of the t.co URL shortener. Over time, Twitter gradually migrated all URL shortening to the t.co shortener.

“Now, whenever you click on a tweet, Twitter routes the URL through the t.co shortener,” said Tom Critchlow, VP-operations for Distilled NYC, an SEO and online marketing company. “So it shows up in analytics as a Twitter referral.”

This is true for all URLs that go through Twitter, even ones that have been shortened with older applications like bitly. According to Critchlow, third-party application users aren't even aware of this. There's no indication that your URLs are running from TweetDeck through bitly, for example, then through t.co and finally through Twitter.

The result of this change has been a more accurate picture of how many referrals are coming to any website through Twitter and a de-emphasis on direct traffic.

“For content marketers, direct traffic was reduced and Twitter traffic went up,” Pollitt said.

Although third-party Twitter applications were probably the biggest culprit in this underreporting, Pollitt said, there are probably other misrepresentations in the average analytics report. Links in third-party mail clients, for example, can also be reported as direct instead, and RSS traffic can show up any number of ways.

“It can show up as direct traffic, too,” he said. “For example, if you use Outlook and click on a link from your RSS burner, you'll have no idea where it actually came from. So it will show up as direct traffic.”

Ultimately, Pollitt said, marketers have to work with the data they have.

“It's a misconception that inbound traffic over the Internet is 100% trackable,” he said. “It's not; it's just more trackable than the older print methods.”