Google Analytics “Direct” Traffic is Lying to You
On one hand, seeing more “Direct” traffic is a good thing, right? It implies that people know your brand so well that they just typed in your URL or clicked a bookmark to get to your site. Sweet, right? This means these kinds of things:
- Word of mouth is happening
- Visitors are returning
- People are bookmarking me
- My brand is gaining familiarity
This is all fantastic if it weren’t for the simple fact that:
Google Analytics is a freaking dirty lying liar when it comes to “Direct” traffic.
From a web analyst’s perspective, “Direct” is the worst. It offers little to no insight on how the visitor found you. But, in reality, not even close to all traffic that registers as “Direct” is truly direct. In fact, one experiment by Gene McKenna from Groupon shows that nearly 60% of traffic showing in Analytics as “Direct” is in fact organic search!
There are several things that cause traffic to lose it’s referring source along the way, and be lumped in with “Direct”. Some are just out of our control and unavoidable. The two major causes of this that I’ve discovered are Hypertext Transfer Protocol Spec 15.1.3 and incorrect or missing tagging.
Hypertext Transfer Protocol Spec 15.1.3
This spec states:
“Clients SHOULD NOT include a Referer header field in a (non-secure) HTTP request if the referring page was transferred with a secure protocol.”
In other words, if a link from a secure page (https://) is linking to a non secure page (http://), referrer data shouldn’t be passed along. So the referrer data gets stripped back down to the “Direct” bucket.
Incorrect or Missing Tagging
Basically any time there’s a link coming in without any parameters from things like email, instant messengers, or social media posing dashboards, referrer data can get lost.
Tagging all incoming links that you have control over with the proper UTM parameters will reduce a lot of these. Make sure to communicate with your entire organization that any inbound links from emails, display ads, social media, etc. have uniform UTM parameters.
What We Did
Apart from the slight SEO benefit offered by securing your site, it’s an inevitable step, just as going mobile is. At my day job, we made the decision now is better than ever to convert the site from http to https. It’s quite the project that I’ll probably write about later, but it’s definitely going to help with the false “Direct” caused by Hypertext Transfer Protocol issues, because now, any links coming from a secure page can pass referrer data on to our pages, because they are secure.
Additionally, I implemented a tagging policy for our social media coordinator to ensure all incoming links from social posts are properly tagged. I also identified a conflict between Aweber’s link tracking and our UTM parameters, and disabled Aweber’s tracking to fix the tagging issue there.
After converting the site to https and implementing this stringent tagging campaign, we’ve still only been able to reduce the number of traffic registering as “Direct” by about 5%. Grrr…
I see it only improving over time as some of those old untagged links fall out of use and all the https links get indexed. Of course, all of this won’t go away, because a big chunk of it is legitimately “Direct”. But we’re on the road to trimming that down to a number closer to the real one.
Even the biggest tagging Nazi can’t prevent 100% of false “Direct”. But with a little discipline and a plan, you can reduce this number and get more accurate, useful analytical data to make decisions off of.
Do you have any tricks for reducing your false “Direct”? I’d love to hear from you. Tweet at me or comment here.