I noticed recently that I was getting an unnatural spike in direct traffic in Google Analytics across several accounts. To get a sudden uplift in direct traffic is quite unnatural even perhaps if there is offline marketing activity going on.
Here’s a snapshot of direct traffic from one account in the past 6 months (where you can see the uplift in May 2015):
We often see lots of spammy referral traffic but to see an uplift in direct traffic meant I needed to dig down a little further.
At first I looked at direct traffic by country and the uplift was mainly coming from the United States. But blocking traffic from the United States probably isn’t the best option for you if your business operates there.
After searching a bit more on the web it seemed people were talking about IE7 being the culprit but I couldn’t really see anything with IE7 either.
But when looking at what browser the user was using in direct traffic, I noticed there was a large amount of traffic coming from Google Chrome. Maybe Chrome would be around a 30% browser share, but this time it was more like 70%.
You might see in the reports that if you go to “Audience>Technology>Browser & OS>Chrome” you see a similar pattern:
I then plotted the rows of the first result in the table (40.0.2214.111) to see if this roughly followed the line (and as you can see, it did):
So the culprit was Google Chrome and Browser Version 40.0.2214.111.
I don’t know why in particular this is. There could well be something that is trying to exploit this version. Either way we want to get rid of it to stop it affecting the data.
To get rid of this unwanted data (and historical data too) I use segments.
To do this click on “+ Add Segment” and then “+ New Segment” in the report view.
Click on “Conditions” in the left hand tab under Advanced.
Now you want “Sessions” and “Exclude” in the Filter along with “Browser Version” and “Contains”, followed by the number “40.0.2214.111” in the field. Like this:
Then click save. This will get rid of the unwanted and unnatural direct traffic skewing your data!
Why SEO is vital to Web Design: 13 Reasons (using a Mexican restaurant)