While reviewing natural search performance for several clients in May, an interesting and concerning trend became apparent – organic search traffic is trending downward.
Or is it?
In SEO, our first response when we see a traffic impact across multiple properties is to scour industry news for rumor of an algorithm update. Though often times, it’s difficult to come across any confirmed updates by Google. Furthermore, speculation based on search engine results page (SERP) volatility doesn’t always paint a clear picture of what may be happening.
So what do we do when nothing out there suggests an algorithm update? We dive into analytics. In SEO (and digital marketing in general), our ability to illustrate the success of campaigns and SEO implementations relies heavily on our reporting capabilities.
Segmenting by channel, device, and referrer is critical, especially as changes to the SERP (read monetization) continue to impact natural search performance, which already saw YoY negative growth across all devices in Q1.
Dark search strikes again
We know from previous studies, such as Groupon’s bold experiment, that Natural search is often underreported for a variety of reasons. Dark search is an increasingly painful thorn in the side of SEOs.
Generally, dark search is thought of as a percentage organic traffic being reported as direct, but this term can also apply to referring sites or “other” traffic as well. Our efforts are becoming more and more difficult to quantify because of disappearing data.
On April 27, 2016, Google increased the amount of “dark search” traffic with an update to the Google App for Android. With the roll-out of this update, we noticed an immediate spike in referrer traffic that coincided with the immediate decline in organic traffic.
The App sends the referrer id of com.google.android.googlequicksearchbox, which had no attributed visits prior to the update. As user devices continue to update to the new version of the App, referral sessions continue to increase over time.
It’s a convincing argument when you look at the combined data for this specific client – approximately 11% of organic mobile traffic from Google has shifted to referral traffic since the App update occurred.
This percentage varies by client, likely due to user demographics, but the shift appears to be present across all clients reporting this referrer. Values range from approximately 5% to 11% of organic mobile traffic from a sample of clients in varying verticals (including MerkleInc.com).
To complicate matters, reporting of referral sessions from com.google.android.googlequicksearchbox seems to be very inconsistent across analytics platforms. There are instances where one platform (such as GA) is reporting a substantial number of referral visits, while another (Adobe Omniture) is reporting none at all. An evaluation of server logs may be the only surefire way to quantify how much traffic is originating from the Google app.
What Next?
Ultimately, this looks to be part of the pattern of incomplete data for organic search. We, as SEOs, must be diligent in our efforts to account for as much natural search traffic as possible in order to paint a clear picture of natural search performance. As we wait to see how this evolves, it is important to consider Google app traffic when reporting SEO performance.
Additionally, we need to keep an eye on referring sites data so we can detect additional cases of traffic shifting from organic. Obviously, the traffic shift discussed here impacts a very specific segment of traffic – Android Google App users, but likely foreshadows what we can expect with the increasing number of “search” technologies like Google Assistant, the Gboard, Spotlight Search for iPhones (which reports as direct traffic), Amazon Echo, Search Apps, and others.
