It seems that whenever I speak at a conference or give a webinar on Google Analytics, someone asks a good question, that prompts a blog post. That happened again recently, when the question was about the high measure of Direct traffic to the property’s website, asking about high direct traffic sessions.
1. Verify the Analytics Code on Your Site
Visit your own website, then right-click on the page somewhere, and select “View Page Source” (or similar language – depending on the browser and operating system you are using). Look for the analytics code. It is usually near the top of the page (though some sites that have not been updated may have it near the bottom). To find it press Ctrl-F on PC (Cmd-F on Mac) and put ga.js in the search box. If you don’t find anything, try again, putting analytics.js in the search box.
Compare every character, as a single difference can cause a problem. There is no problem if your code has an entry saying “_gaq.push([‘_setAllowLinker’,true],[‘_setDomainName’,’YOURDOMAIN.com’]” or something very close to this.
2. Are You Using a Redirect Page?
A redirect page is a page that forwards a visitor from one page to another. Some properties may have outside sites (like B&B directories) point to a redirect page, which will forward them to the home page. In that case, GA would count that as direct traffic (or possibly as a referral from your own site). It is possible to do that kind of redirect properly to work with GA, but it is very difficult, and it is very rare to see it set up correctly.
You can check for a redirect by visiting your listings, on directories, TripAdvisor, etc. Go into your listing and see what the links on that site pointing to your site are. If the link is directly to your home page (or another valid page on your site), everything is fine. If it goes to a page that actually forwards the visitor to another page, that could well be the cause of high numbers for direct traffic.
3. The Biggest Cause
The biggest cause of (valid) Direct referrals these days is people coming to your site from a mobile app. If you are listed on an app, apps do not send referral information (or, at least, it is very rare for them to do so), so Google lists these as Direct.
One way to find out about this is to first go to the Audience menu in Google Analytics, then Mobile > Overview. Notice the percents of visitors using desktop, tablet and mobile. Then go the Acquisition menu > All Traffic, then click on the (direct) link. Next, under “Secondary Dimension” select Operating System. This will sort your Direct traffic by operating system.
What you’re looking for is the “desktop” operating systems, compared with the mobile systems. Mobile would be Android or iOS. If those seem very high (high is relative to the amount of overall mobile traffic on your site from the first area), then your site is being reached by mobile devices without a referral source. If you have any control over the links these apps use to refer traffic, there are things that can be done to improve it, but they are far from foolproof. Of course, as mobile apps continue to be used, that problem will not go away.
If your mobile traffic is in proper proportion, then it isn’t apps that are creating the Direct traffic. One other thing to check is to go to Audience > Technology > Browser & OS. Then add the Secondary Dimension of Source. Sort the result by clicking on the Source title of the column, then re-sort by clicking it again, so the (direct) entries are at the top. What you’re looking for here is whether there is any single browser version that is causing a very high proportion of the Direct traffic. It has occasionally been the case in the past that a particular web browser had default settings that caused a high proportion of Direct traffic.
Like many issues with traffic, and analysis with an analytics tool, like Google Analytics, the tool gives information. It is up to you to draw the right conclusions from the information it gives. Sometimes it gives very strong clues. Sometimes the clues may not be clear enough to justify a conclusion.