Recent Facebook announcements about upcoming News Feed changes have been underscored by widespread sharing of the story that publishers are seeing declines of 20-something% in organic Facebook reach already over the last year. Digging into the data a little deeper, I think the story is even worse: I believe publishers are seeing traffic outside the Facebook ecosystem down more like 40%.
The last few years has seen breathless hot-takes, and dramatic over-reactions to the ebbs and flows of organic Facebook reach and traffic.
Off the top of my head, I remember:
- Buzzfeed reporting Facebook referral traffic outstripping Google traffic back in 2014 and it being covered with such moderate, careful headlines as and just like that Facebook became the most important entity in web journalism (The Atlantic) and the year Facebook blew past Google (recode). The latter failing even to make it clear that search was certainly still bigger outside of the media bubble
- DefineMG responding that outside the Buzzfeed network, Google was still well in the lead in 2014
- Parse.ly reporting in 2015 that a broader set of media sites had seen Facebook overtake Google (43% to 38%) - again with too many people forgetting that this was only about media companies
- Research by Moz and jump shot showing that Google is absolutely dominant in overall referral traffic across the whole web (63% to 6% for Google and Facebook respectively)
Then, throughout 2017, we saw a pretty steep trend downwards in organic visibility on Facebook causing big drops in Facebook’s referral traffic out to other sites. As if this wasn’t enough writing on the wall, Zuckerberg kicked off 2018 by announcing major news feed changes that will dampen organic visibility of media sites even further.
This was entirely predictable - we have been speculating that Facebook could even remove links entirely from the news feed (much like Instagram) if they lead outside the FB ecosystem. Nonetheless, it’s caused much media hand-wringing [warning: possibly NSFW headline].
The reporting of declining Facebook referral traffic through 2017
Over the last few years, Parse.ly, which provides analytics tools primarily to publishers, has repeatedly published data showing the ebbs and flows (I linked to some of their data above). No doubt they’ll continue to do this and show us what the 2018 news feed updates look like for media.
It’s been interesting to me though that the charts they have been releasing, and that have been widely covered have been presented either:
- as a % of total referral traffic (effectively Google+Facebook) which looks like Google is growing strongly when in fact it’s only showing modest growth primarily in AMP; and/or
- including AMP/instant articles which obfuscate some of what I think is happening
When you look at a chart like this widely-shared one:
It’s basically impossible to tell what’s really going on with Facebook referrals - because what is really happening is a much steeper Facebook decline (this is a % chart and the total is declining for this sample set, driven by the Facebook drop.
Facebook referral traffic to normal web pages is down closer to 40%
I think that the more interesting chart is actually one of the less-shared ones in that Parse.ly report. This chart shows absolute traffic levels rather than %s, and if you read it carefully you can tease out four numbers - Facebook traffic to instant articles and to regular pages and Google traffic to AMP and to regular pages:
By reading the chart carefully, and working with the % numbers shared in the report itself, we get the following changes:
Google traffic changes for Parse.ly publishers in the last year
- Google AMP pages: up 87%
- Google non-AMP pages: up 6%
- Google total: up 17%
Facebook traffic changes for Parse.ly publishers in the last year
- Facebook instant articles: up 3%
- Facebook total: down 25%
So far, so much as reported.
But for me, the mind-blowing implication is that if instant article traffic is basically flat, but total Facebook traffic is down 25%, that means that Facebook traffic to the normal web pages in the Parse.ly network is down 38%:
I think this is the more interesting number because it is the metric for traffic outside the Facebook ecosystem. It’s great that Instant Articles are still doing OK (though +3% pales in comparison to Google AMP traffic up 87%) but the issues with Instant Articles and the fact that it’s a clear path to everything staying entirely within the Facebook ecosystem would have me very concerned to build a business model on it.
I’ve been fascinated at the surprise that I’ve seen from sharing the Parse.ly data - a 25% decline is a big deal, but I had assumed it was bigger than that for many publishers, and digging into the data appears to back me up.
If you want anecdote to go with the data, I enjoyed listening to this podcast episode from Recode Media that describes the cratering of Facebook traffic to Bleacher Report:
Our traffic from Facebook is a fraction of what it was at the peak. I’m talking like 15, 20 percent of what it was.
While I’m not sold on the “pivot to Instagram” as a long-term strategy (fool me once?) I’m rooting for BR and am fascinated to see where it goes.
Meanwhile, all of this is far more dramatic than was generally reported in coverage of the parse.ly report, and remember that it is before the 2018 news feed changes kick in so we are likely to see plenty more of it this year. It’s not hard to predict the 2018 trend, so I’m interested to see the various responses we see from different publishers. If you’ve seen public statements on their plans, please do share them in the comments - I would definitely be interested to see them.