If you're seeing a rise in direct traffic in Google Analytics and wondering what could be causing it, there are a few common culprits. As of late though, SEOs and online marketers have been dealing with what's called Dark Traffic. The point of this guide is to diagnose your direct traffic and see if it's actually NOT direct.
The first step is to first rule out everything besides Dark Traffic. We want to avoid confirmation bias and look at our metrics with a clear head. A few preliminary questions to ask before actually looking for Dark Traffic are:
Do I have bot filtering on?
If yes, then move to the next question
If no, then go turn it on. It can be found in Google Analytics under Admin > View Settings > Bot Filtering
Is my internal IP filtered?
If yes, then move to the next question
If no, then create a new view and filter out your IP address
Are my web developers testing website availability by using a ping bot?
If yes, then that is most likely your culprit
If no, then let's start looking for Dark Traffic
As mentioned in my Dark Search blog post, Dark Traffic is a URL that cannot be tracked because it does not pass along a referrer string. This means Google Analytics cannot determine a traffic source, and is categorized as "Direct". It's possible that this traffic actually isn't direct, but is referral, social, or even organic, but don't assume that it is. We have to prove that it is indeed Dark Traffic. Here's how:
Pull your direct traffic: Dark Traffic is only categorized as direct. You won't find it anywhere else. This can be done by going to Google Analytics, then Acquisition > All Traffic > Channels > Direct (click on it under the Default Channel Grouping column). Now you are looking at only traffic categorized as direct.
Filter out pages that actually could have direct traffic: This can be pages that are commonly visited or bookmarked. A safe bet is to filter out your home page and section fronts. This means links in your navigation menu. To do so:
Click on "advanced" near the search bar
Add a Dimension
Select Landing Page
Select Exactly Matching
Add the page you will be filtering out
Repeat for each page
Dark Social: Now that you have removed your home page and section fronts, what's left is called Dark Social. Dark Social is one of the three major buckets of Dark Traffic. You'll notice that the URLs here are longer in length and may include multiple directories. These are the types of URLs that have a lower chance of being actual direct traffic, but we are still not 100% sure.
Eliminating Dark Social: To filter out possible traffic sources even further, we need to cross-reference our social campaign with the most popular emails. Reach out to your social media manager or team to see if any campaigns were pointing to these URLs. If so, then you have found the actual source of the traffic, and your social media manager can call out this traffic in their reporting. Next, filter out those URLs that had social campaigns running.
New Users: After you have eliminated any traffic that may be from social campaigns, we need to eliminate users that may have bookmarked these URLs or reached them via the address bar auto-complete feature. Do this by clicking on New Users under the Acquisition column.
Dark Search: Now what you have left is Dark Search, traffic that reached deep links on your site from an unknown, non-social source. This is traffic that can be organic or referral traffic with a referrer string stripped.
Possible causes: Unfortunately, from here we can only investigate further to find actual sources of traffic, or just leave it be. Some common causes of Dark Search are: HTTPS referral links to a non HTTPS site, email links without UTM sources, mobile traffic from apps, chat services, or improperly tagged paid campaigns. Look into some of these possible channels to get a better idea of where your traffic is coming from, and you could find a potentially un-explored source of traffic.
Steps to cut down on Dark Traffic:
Tag all of your off-site marketing initiatives with proper custom links and UTM markup
Suggest switching to HTTPS
Look for correlations in marketing campaigns before reporting KPIs