Spike in direct traffic? It could be Google Discover

I work on business news websites, and a little while back I started seeing the occasional large spike in 'direct' traffic in the Google Analytics accounts for these sites.

The traffic spikes were always to a recently published news story, and tended to range from 2,000 to 20,000 sessions. I wanted to get to the bottom of what was driving this significant level of direct traffic. I had two aims: to prove that it was genuine traffic, and to see whether I could capitalise on the source and potentially gain even more visitors from it.

Defining direct traffic

But where does direct traffic come from? Google Analytics somewhat amusingly describes direct traffic as traffic from "users that typed your URL directly into their browser, or who had bookmarked your site". This is like saying that fruit is an apple or a pear - not exactly wrong, but hardly the full picture.

I also find it impossible to believe that users typing a URL directly into a browser or bookmarking the site would be responsible for a spike of 20,000 sessions.

The definition that SEO software company Moz gives in its Complete Guide to Direct Traffic in Google Analytics is more comprehensive - and useful:

Google Analytics will report a traffic source of "direct" when it has no data on how the session arrived at your website, or when the referring source has been configured to be ignored. You can think of direct as GA’s fall-back option for when its processing logic has failed to attribute a session to a particular source.

So in effect, direct traffic could have any number of causes. In the past I've tracked down a lot of direct traffic as being the result of an https > http > https redirect chain. However this tends to cause a steady flow of direct traffic, whereas the spikes I was seeing now were very short-lived - each occurring over a maximum of three days (with one day much higher than the other/s). So I didn't think that redirect chains were the culprit.

However it wasn't until one site received a spike of 130,000 sessions that a colleague managed to identify the source: Google Discover (previously known as Google Feed). Search Console Help has a really nice explanation of what Google Discover is and does, so I won't repeat that here. Only to add that I can personally vouch for the truth of the following:

Given the serendipitous nature of Discover, traffic from Discover is less predictable or dependable when compared to Search, and should be considered supplemental to your Search traffic

Uncovering Google Discover

But how did my colleague know that Google Discover was the cause? The figures in the Google Search Console Discover Performance Report (AKA Performance Report for Discover!) tallied closely with what we were seeing in Google Analytics.

The report itself is found under Performance > Discover, but note that it "is visible only if your property reaches a threshold number of Discover impressions in the past 16 months." So if you cannot access this report in Search Console, then you can be fairly confident that Discover is not the cause of your traffic spikes.

If you can access the report, here is how you might compare the data within it to your Google Analytics data.

  1. Log in to Google Analytics
  2. Go to Acquisition > All Traffic > Channels
  3. Click on Direct
  4. Your spiking story should be listed here under the dimension 'Landing Page'. Note the number of sessions
  5. Log in to Google Search Console
  6. Go to Performance > Discover
  7. Find the same story and note the number of Google Discover clicks
Google Discover Performance Report

Part of the Discover Performance Report in Google Search Console. Clicks and impressions are both very spiky - and average click-through rate (CTR) is high

There is of course not a 1:1 relationship between Google Discover clicks and direct sessions. Direct sessions have other causes too, so this figure is likely to be slightly higher than the number of Google Discover clicks. For example you might see:

Google Analytics Google Search Console
Story 1 24,573 sessions 23,823 clicks
Story 2 10,508 sessions 10,202 clicks

If the two figures for any given story are very similar like this, then you know that Google Discover is responsible for the majority of direct sessions for this story and hence the spike. Incidentally the figures I obtained also suggested that all Discover traffic is classified as direct - at least in the case of the sites I looked at.

It would be nice, I think, to build a dashboard in Google Data Studio that included both Google Discover data from Search Console and direct acquisition data from Google Analytics. This would enable you to see at a glance whether a direct traffic spike was the result of Google Discover.

Subscribe
Notify of
guest
0 Comments
Inline Feedbacks
View all comments
James Clark
Hi! I'm James Clark and this is my website. Need help with your own website's analytics, SEO, content or ads? Contact me
Copyright © James Clark
0
What do you think? Leave a commentx
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram