How Traffic Moved Across The Web Q4/16
The dynamics of content discovery are changing. From day one of the web, through today and likely always, getting people to your website from another website…and then sending them to someone else’s website, is the most important part of being successful on the web.
At Yieldbot, our technology sits on publisher websites, enabling us to understand the real-time intent of consumers. We focus our learning on what people are clicking on. Our first step is discerning where a person came from to start their session. Each source of a visitor to a site is a unique data segment into understanding their real-time context, a large part of understanding relevance and intent.
The dataset below covers Q4, likely the most talked about quarter relating to content and the web in its history. According to comScore, Yieldbot covers 93% of the U.S. online population and 204M unique people a month, making this one of the largest data sets of its kind — over 15 billion referred sessions to over 1000 publishers across diverse verticals.
Some important notes:
- There are many more referrers than represented here. It is a long list with hundreds of millions additional referrers, but as single sources they are not consequential. Arguably nobody is, except Google and Facebook.
- Facebook sends four different referrer sources based on m. (mobile) l. and lm. (LinkShim a system used to protect malicious links) plus their regular domain covering desktop and laptops users. Combined, it’s safe to say Facebook is now the largest measured referrer source on the web.
- Twitter’s link redirect is t.co. The traffic they drive to publishers is less than 1% of referrerals. Tells you quite a lot about Twitter.
- Outbrain distinguishes between its ads and content referrers while Taboola does not.
- Bing is a significantly larger source of visitors than Yahoo Search.
- AMP, while a big Google initiative did not send a huge amount of referrerals. It is very important to realize that Google caches the publisher pages in AMP so there are large, unknown amounts of organic searches that never reach the publisher and stay within Google’s domain.
AMP (aka Google Accelerated Pages) may be the single biggest reason Facebook is now surpassing Google in traffic. One way to think about this is if the CTR from AMP to a publisher is 10% then there were 330M more referrerals that never got to the publishers own sites. Likely that number is much higher. There is a big story about publisher monetization (or lack thereof) in AMP, but that’s for another time.
The biggest story is the rise of Facebook and thus mobile. Facebook is authenticated traffic and performs very well across back-end metrics for publishers and brands. Having two giant sources of traffic is a great thing for publishers. It gives them more options and brings consumers to their site with different mindsets (social vs. search). Interestingly one of these giants is playing nicer with pubs than the other.
The other big story is that a huge percentage of traffic is blind to publishers. It is imperative that publishers work on ways with their web analytics to deduce the traffic sources that make up direct traffic and attribute it as best they can.
Not understanding site visitors just cedes more advantage to buyers and their profile data. Understanding your visitors should be a key role in any organization that sells something. It is impossible for publishers to develop relevant brand building experiences without understanding the context of how their audience is navigating to their site. This way referrer data is so important.