Most serious Internet marketers look at referred traffic and direct traffic data from their websites. But do they really know exactly what those traffic numbers represent? There is much more than meets the eye in terms of how Google Analytics buckets referred and direct traffic, and in addition, a number of issues can skew the numbers in ways that make it difficult for companies to accurately assess the effectiveness of their campaigns.
In this technical article, we’ll bring clarity to what the numbers mean — and how you can make the numbers more meaningful.
1. Direct Traffic Composition
Direct traffic is made up of visitors who reach a website without a referral URL. Here are some examples of traffic that will result in a direct source:
- Typing in a website URL in your browser’s address bar
- Clicking a bookmark in your browser
- Clicking a link in a document (Word, Excel, PowerPoint, PDF, etc.)
- Clicking a link in an application (Outlook, Mail, etc.)
- Clicking a link from an instant message
- Clicking an HTTP link from an HTTPS website
The Google Analytics cookie will track the referring source of all visits you make to a website. By default, Google Analytics is set to a last non-direct click attribution model, which means the marketing channel responsible for the last non-direct click gets the credit. So, if your first visit was from Google organic, your second visit was from referral and your last visit was from direct, Google would credit Referral as the source. However, for multi-channel funnel reports, Google Analytics defaults to the first interaction model. Google Analytics has a processing flow chart that describes how the marketing source is determined.
2. Direct Traffic and HTTPS
In August 2014, Google announced that it would start to include HTTPS as an organic ranking signal. This announcement started a migration of webmasters to begin moving their websites to run on HTTPS. Prior to this, HTTPS was rarely used outside of e-commerce checkout pages and login pages protecting sensitive areas of websites. As more websites move to HTTPS, direct traffic will increase due to visitors clicking HTTP links from HTTPS websites. Normally, this traffic would have been tracked back to the referral site. Not having this referral site information is an issue for online marketing. There are a few ways for a website to get around this issue:
- Modify your web server to rewrite the referral URL
- Create a script to rewrite the referral URL
- Implement the meta referrer tag
The issue with these solutions is that the website receiving the referral traffic is the one that cares about fixing the issue, but the website sending the referral traffic is the one that needs to fix it.
3. Direct Traffic Spam
If you receive a lot of direct traffic or see a spike in direct traffic, it is worth investigating to see if you have a direct traffic spam issue. The main cause of this spam traffic is bots. In the old days, bots would not process JavaScript and therefore would not show up in Google Analytics. However, this is no longer true as bots can now process JavaScript. Spammers commonly create evil bots to either attempt to make money or to cause issues with web servers and increase costs for website owners. One study shows that these evil bots account for 29 percent of all website traffic. Here is one way to see if your website has direct traffic spam in Google Analytics:
- Go to Acquisition > All Traffic > Channels > Direct
- Click your home page for the Landing Page
- Set your Secondary Dimension to: Behavior > Page Depth
The first row should show you the number of sessions that reached your home page directly and had a bounce rate of 100 percent. If you see a high number of sessions, it is safe to say that your direct traffic metrics include spam.
Blocking direct traffic spam isn’t easy as it normally comes from multiple IP addresses in different countries, states and cities. It can also come from a variety of different service providers and user agents. Assuming you don’t want to block all traffic from certain countries, states, cities and user agents, there is little that can be done to completely fix the issue. However, there are a few things you can do to reduce spam from being counted in Google Analytics:
- Configure the option under Admin > View > View Settings > Bot Filtering, which will exclude all hits from known bots and spiders. However, we see that this rarely works to combat the traffic spam.
- Examine your web server logs or use a tool like Opentracker to identify IP addresses that are driving spam traffic. Then, block these IP addresses in Google Analytics using view filters. A better way to block IP addresses is to do it at your web server level; if you run Apache, you can do this with a .htaccess file.
- Configure an advanced segment in Google Analytics using conditions to only include traffic to your specific hostname and exclude direct traffic that is specific to outdated browser and flash versions hitting specific pages that you’ve identified as being responsible for the majority of spam.
- If all else fails, you can manually filter out all direct traffic to your home page with a bounce rate of 100 percent.
Without making direct and referred traffic data as accurate as you can, you’ll be using more guesswork than you should to understand how well your Internet marketing campaigns are working. And even if you can’t make the numbers 100 percent accurate, at least you will know what the flaws are, and interpret them in a reasonable and reasonably precise way.