How Facebook’s Newsfeed changes are affecting European publishers
Back in January, Facebook made an announcement that shocked the publishing world: It was tweaking its Newsfeed algorithm so there would be less emphasis on Facebook pages and more focus on what your friends were sharing.
Reach for Facebook pages was expected to plummet, and publishers that had grown addicted to Facebook’s free referral traffic girded themselves for an all-out decimation of their businesses.
But while there have been some clear victims of the algorithm change, the actual data for who’s been hurt by it is noisy. Some publishers have seen a dip in Facebook engagement, while others have remained steady or even improved.
That’s the case for U.S. publishers, at least. But what about Europe? To find out, I interviewed Steve El-Sharawy, head of innovation at EzyInsights, a platform that tracks Facebook engagement for publishers.
I asked El-Sharawy about how European publishers reacted to the announced algorithm changes and whether there are any clear winners and losers in the ordeal.
To listen to the interview, subscribe to The Business of Content on your favorite podcast player, or you can play the YouTube video below. If you scroll down you’ll also find a transcript of the interview.
iTunes/ Stitcher/ Google Play/ Overcast/ Spotify
A transcript is below.
Simon Owens: Hey Steve, thanks for joining us
Steve El-Sharawy: Thank you for having me.
So you’re the head of innovation at a company called Ezyinsights. Can you give us a sense of what your company does?
Absolutely, yes. What we do is we have a tool we sell mainly to publishers — news publishers, content publishers around the world. We discover content and articles as they are published and we look at the engagement on social media that they pick up so we can order them into what people are most interested in at any given moment.
And one of the major platforms you track is Facebook. You’re kind of analogous to a tool a lot of U.S. publishers use called CrowdTangle. And so you’re able to actually track the actual engagement rates for publishers’ Facebook pages, which is immensely insightful for trying to track longterm trends and how Facebook’s algorithm is treating publishers.
That’s correct. We are very similar to CrowdTangle. I suppose the main difference is that CrowdTangle looks specifically at the Facebook posts that are made. Now we look at Facebook posts, of course, but we also look at websites. There’s one particular metric, webshares, where people will take the URL of an article and they’ll then share it onto their own Facebook page or someone else’s Facebook page. We track that as well. I would say that, on average, it accounts for 50 percent of engagement for traditional news publishers, at least.
The fun part is analyzing, historically, and seeing those trends happening.
And your clients are major brands, publishers that utilize Facebook in their marketing strategies.
Right, we have newspapers, broadcasters. Anyone publishing news. We have viral pages, magazines. Anyone who’s publishing content online, really. I could reel off a list of names, but they’d be pretty Euro-focused. Mainly our customer base is in Europe. Some in South America as well. We don’t actually have any customers in the U.S. right now.
As we all know, Facebook announced a major change to its algorithm in January that would downgrade pages and publishers within the algorithm in exchange for upgrading posts from your friends and also Facebook groups. How did your clients react over in Europe to the news? Was there a sense of panic?
I think in general there probably was a sense of panic. It didn’t really filter through to us via our customers. And I think one of the reasons is that our customers tend to be our service, and they tend to be on the ball when it comes to what’s happening. So they don’t have such a reason to panic because it’s not going to be unknown to them. But we did have quite a lot of questions from people who weren’t quite customers yet. Worrying that this would really affect them, that this would kill them. So certainly, a bit of panic, but more, I would say, trepidation and a kind of sense that they just need to keep an eye of what’s going to be happening over the short to medium term a little more closely.
Would you say that the announced change spurred more interest for analytics tools like yours since they could no longer rely on the free engagement from Facebook?
That’s an optimistic way of looking at it. Yes, possibly. I think that’s certainly something that could have happened. I don’t have any statistics in front of me. But it certainly brought in the idea of Facebook engagement and what it means more sharply into focus.
A lot of publishers, ever since this algorithm change was announced, have been claiming to be victims. One group that’s been especially vocal has been conservatives. Here in the U.S. they’ve been claiming all kinds of censorship. Here in the U.S. there were these two pundits, Diamond and Silk, that were interviewed during Congressional hearings. When CEO Mark Zuckerberg was interviewed, he was grilled about this supposed censorship. There’s this widespread conspiracy theory that Facebook has an anti-conservative bias. When you looked at data for right wing Facebook pages, what did you find?
We’ve been tracking what we call controversial pages for quite a long time now. And we saw engagement go down across controversial pages, and across all news publishers. We saw that there was a narrative out there that far right publishers, and maybe controversial pages as well, are being targeted, and bullied down by Facebook. In fact, what we saw on content from those pages, was a lot of posts pleading with their followers and their fans saying hey, you follow us on Facebook but you might not see us in the future so much, so please, subscribe to our newsletter. Or follow us on another platform. Or subscribe to our Whatsapp channel. There was definitely a sense of panic, or maybe an opportunistic sense to drive their readers from Facebook into other channels.
You found that engagement was down for some of these controversial pages, but it was kind of a mixed bag because your report said that part of the reason for the engagement fall was because some of the pages got deleted/banned so it was throwing off the metrics a little bit.
We decided to investigate that and take an unbiased look at the data and see if it’s really true. And we certainly didn’t see anything that would suggest that conservative pages, or any particular kind of pages, were being unfairly targeted. So what we saw was that conservative and far right leaning pages, they do really well. So there isn’t necessarily comparisons between extreme Left and extreme Right pages on Facebook, but we organized them into both sets. But one thing to think is that, on the Right, there’s a huge amount of engagement that’s not only happening now, but has been happening over the last few years. The drop they’ve had — — I think about a third of it was from page deletion. So quite high profile manual page deletions. If you look in the UK, Britain First’s page was finally kicked off for contravening their guidelines multiple times. There are still quite a few high profile ones that are active. But what we saw, in terms of overall decline, even taking into account those manual page deletions, was in line with general trends we’ve seen across all news publishers, be they on the fringes, or be they on the mainstream. We’ve analyzed several different countries. We’re working through Europe at the moment. And so far we’ve seen that the Netherlands have been most affected, in terms of percentage. The news industry there, they’ve dropped slightly more than any of the far right pages on their own.
If I’m understanding correctly, there was so much engagement on the far right pages, that even if they’ve seen a steeper drop than some of the other categories, they’re still, overall, seeing more engagement than most other publisher categories.
Absolutely. That’s exactly right. One thing we noticed looking at this engagement. We looked at the last five years. One thing we saw that was, page creation — so pages that were created on Facebook and began to gather fans, began a lot earlier with conservative pages. So they’ve really been doing this since well before the presidential election of 2016. And a lot of these pages have a lot of followers. So they’ve had a big headstart, in a way. And even though they’ve had this drop, which a lot of publishers on Facebook have had a drop, they’ve still been significantly higher than any other equivalent that might be on the other side of the political spectrum.
I think data reflects that in the U.S. as well. They did some analysis with CrowdTangle an Newswhip with Diamond and Silk and some of these other conservative pages, and yeah, the data is really noisy right now. It seems like the algorithm changes and the effects are ongoing, but overall, at least in the U.S., liberal pages have seen a huge drop in engagement, much steeper than conservative pages. Donald Trump, his Facebook engagement has remained relatively steady, whereas Elizabeth Warren and Bernie Sanders, two liberal firebrands who have done well on social media, their pages have seen really steep drops in engagement. So, at least based on the data I’ve reviewed, I’m not seeing any specific targeting of conservatives in general. Obviously they’re a little bit more likely to step over the line and violate terms of service, and as you noted in your report, they’re a little more likely to be deleted overall, but in terms of the pages that stay up, it doesn’t seem like the engagement is out of line with the drop in engagement with other publishers.
No, that’s a good observation. When we tried to find a trend, find some common elements between the content that has fallen the most, we do see a bigger drop when it comes to tabloid-type content. Sometimes we call them engagement catchers. Posts that encourage people to perform some kind of action, empty engagement that drives the numbers for tabloid publishers. And they’ve been getting used to these peaks which drives their reach. It seems Facebook has targeted that kind of engagement. That kind of engagement works in the light entertainment industry, celebrity pages and things like that. So we’ve seen quite a bit of drop on their pages. But it’s also a method that’s quite commonly on the far Right. They’ll use it in a slightly different way, but they’ll encourage people to engage in that way. And whereas, the difference is, on the liberal and far Left pages, we don’t see quite the same approach when it comes to their content. We see them much more sharing stories from the mainstream media. From traditional publishers. Just asking people to read it and click through, and not asking them to do something specific. So we see that particular type of content is targeted everywhere, not just conservative types of pages.
[LIKE THIS ARTICLE SO FAR? THEN YOU’LL REALLY WANT TO SIGN UP FOR MY NEWSLETTER. IT’S DELIVERED ONCE A WEEK AND PACKED WITH MY TECH AND MEDIA ANALYSIS, STUFF YOU WON’T FIND ANYWHERE ELSE ON THE WEB. SUBSCRIBE OVER HERE]
Maybe this is my own personal bias stepping in, but I think of liberalism being a little more big tent ideology with more diverse interests, whereas the conservative ideology is more in lockstep, so it’s probably harder to get the liberal community to act in unison where they’re showing huge waves of engagement for every single cause, whereas right wing groups it’s a little more tribal. And so, you did a specific study for Facebook pages for publishers in the Netherlands. Overall, how much of a drop of engagement did they see?
The Netherlands was quite an alarming case. We have quite a few customers there. We really saw a lot of them dropping in terms of Facebook engagement. We took a six-month period, from about May this year and previous to that, and we saw a drop on the top publishers by about 36 percent, overall. It wasn’t very good news for them. One thing we did see was quite a few local news publishers and local broadcasters. Really quite small pages did go against the grain and went up in their engagement. I’m not sure the engagement they gained was enough to offset the engagement that was lost. I shouldn’t think so, because those pages were quite small. However, the Netherlands is an interesting country the way the media is spread out across the country, because they have tons and tons of publicly funded, local broadcasters and publishers. So if the Facebook algorithm changes were meant to boost local news and bring down the bigger publishers, it would have a bigger effect in the Netherlands where there are a lot of these smaller local publishers.
Facebook did announce that they’re trying to place more emphasis and that they’re rolling out all kinds of new tools to help local publishers. So that’s interesting that you’re seeing that effect. In fact, you’re the first place I’ve heard actual reporting showing that local publishers are seeing an engagement boost. So that sounds somewhat encouraging.
It is encouraging and it’s not encouraging. The problem is, when you look at this kind of thing at a country level, and you go to the country level, and you look at the resources that are behind broadcasts, they’re different across all different countries, and they’re different from area to area. In some cases they’re publicly funded and other places commercial broadcasters do well. It’s all very well to say we want local pages to do better in this country, it’s tampering with an entire news ecosystem that exists from country to country. It’s kind of a blunt way to do it, to tweak an algorithm and say, we decide who does better. In the Netherlands, a lot of the local broadcasters are publicly funded. So what you’re getting is the publicly funded broadcasters are doing maybe slightly better at the expense of commercial broadcasters, who are the national ones. It creates more of a tension between these publicly funded broadcasters and the commercial ones who complain that the public ones are competing with them and taking engagement way.
So you saw in this report a huge drop in engagement for video. It looked like the posts for links and photos, they saw a little bit of a drop but remained relatively steady. So a lot of the drop in engagement can be attributed to video. Why do you think that is?
That is exactly what we saw. We saw video drop the most. We saw links, articles posted on Facebook, they remained relatively steady. And interestingly we saw photo engagement go up. Which hasn’t gone up in a long time. So as for why, it’s a difficult question, because Facebook has been pushing video for such a long time and encouraging publishers and broadcasters to put more native video on Facebook, and it’s a resource heavy medium. It costs money to set up a studio to go and put out quality content and put it to your video. And it takes traffic away. If it’s native video on Facebook, it’s just providing traffic on Facebook. And then all of a sudden, we see a drop. So it’s a little bit of pulling the mat from out underneath someone is the feeling that you have when you see this video engagement drop.
And Facebook has been all over the place with it. First they rolled out Facebook live and they were paying publishers to create live video. Then they pulled back on that. Then regular video seemed to be doing well. And now they launched the Facebook Watch tab and they’re throwing upwards of a billion dollars in the next year producing shows for that. Perhaps this is part of that kind of shift now they’re focusing more on Facebook Watch than native video in the feed. Videos in the feed are seeing a drop in engagement. I’ve also seen that publishers in general have been shifting to YouTube because they find that Facebook video engagement is really shallow, especially since Facebook counts a view as only three seconds versus 30 seconds for YouTube. It seems like there’s been a big shift away for publishers from Facebook video in exchange for other platforms like Twitter video and YouTube. It seems like Facebook’s been a little schizophrenic in how it’s treated video in the last year or two.
I would agree with all of those points. Interestingly, maybe one thing we saw when Facebook first introduced Facebook video, is that even though their official statement was that we’re not giving it any preferential treatment within the Newsfeed, we saw the entire news landscape completely dominated by native video, so what happened was, the publishers that could react the quickest to that changed and optimized for that, and they got huge boosts. Sites like Ladbible, they really took advantage of that wave. And although they tuned it down after awhile, we saw links and articles get into the top 20 posts on any given day, we saw a slight rebalancing. There could be an argument that by bringing the importance of video down is more of a rebalancing to make a more balanced landscape there for different types of content on Facebook.
What strategic advice are you giving to clients who are noticing a dip in their Facebook engagement?
It really depends on the type of publisher that we have. So news publishers, in general, 80 percent of the posts they’ll create are links, so there’s not very much for them to change. There will be things they might like to tweak, might like to be able to improve. And we can point to publishers and say they’re doing this, and this is really working. But essentially, it’s the same as ever. And what that means is they’re looking at what works for them. They’re looking at what works for everyone around them. And if they see something that’s particularly effective, whether it’s Facebook live or something much smaller, like making sure your caption is put in a certain way, or making sure not to ask for cheap engagement, they’ll just be monitoring that, and seeing whether they should make tweaks, and what kinds of stories work well. Analyzing in the short term, and the medium term as well. So looking back historically and seeing hey, we have good stories, there’s a big theme here, and the big theme is public interest, which is slightly different from what our most popular stories are on our website. So let’s put more of these kinds of stories when we have them.
So because they can see what’s trending on Facebook using your tool, they can keep optimizing what kind of content they’re producing so it fits that flow.
Right. And I think it’s a very easy thing to say, hey all you have to do is monitor the trends and data coming out, and technically it is. With tools like ours and Crowdtangle and Newswhip, there are more publishers than ever who are plugging into this data. However what we see, is that the top publishers, the biggest ones, the most advanced ones, they are doing exactly what you said. But there is still a cultural hurdle among many news rooms. They have traditional journalists in there, they still are wary of data. They don’t quite understand what we mean when we say your engagement is up or down. And so there’s an ongoing challenge internally in many newsrooms across Europe about how we can turn on our journalists and our newsrooms to understand this data. And to help them.
What’s the overall view of Facebook for publishers in Europe compared to what you’ve seen with U.S. publishers.
It varies from country to country. So in the Nordics — we’re based in Finland — I would say in Finland, Norway, and Sweden especially, they’re very digitally minded. If you look at the biggest tabloid in Sweden, more than 50 percent of their revenue comes from digital, so they’re really focused. They’re not waiting to see what’s happening with social like some of the more traditional publishers could be. They are really forward thinking. They know they really need to use Facebook. They know Facebook is a tool currently, but they don’t make any decision like hey, we’re going to stake everything on Facebook video. And then when we get tripped up we’ll blame Facebook. They have considered this idea of what would happen if Facebook stopped being a news platform. What would we do? And they have systems in place. They have apps, they have strong websites. They maintain all that, as well as optimizing for the Facebook platform. We have each of these different platforms, whether they’re viewed positively or negatively is almost irrelevant, instead it’s based on whether they’re bringing traffic. We’ve had discussion around Facebook how publishers feel. But if it’s bringing in traffic, then it’s considered in an important way. And I’ve certainly seen a shift in attitudes in the last two or three years, where most journalists would be on Twitter. And unfortunately in the Nordics, Twitter isn’t necessarily as big as it is in the U.S. All the journalists are on Twitter. The idea of it are bigger than the reality. We have lots of journalists who say I keep an eye on Twitter all the time. And then I would say about 18 months ago, I started seeing presentations from journalists saying hey, we used to keep an eye on Twitter all the time, we still do, but it only provides 2 to 4 percent of our traffic, whereas Facebook provides 20 to 40 percent of our traffic. So if it’s providing that much traffic, then it needs to be considered with that percentage of importance. Whether they like it or not really.
Are European publishers more open to large platforms like Google and Facebook being more strictly regulated?
I would say almost certainly yes. There’s a lot of friction when it comes to having platforms out of Europe making content decisions. Which we’ve seen when particular pictures were being deleted for running afoul of Facebook’s U.S.-based community standards, which are not the same in different areas of Europe, even from the south to the north of Europe there are different standards and norms that apply. We had the Norwegian prime minister share something and have her Facebook page banned for a little bit. That kind of thing is certainly a cause for concern amongst European publishers and broadcasters.
- EzyInsights gathers data from publishers around the world. Research results cited in this interview were produced by Data Scientist Varpu Rantala.
Did you like this article? Do you want me to create awesome content like this for you? Go here to learn how you can hire me.
Simon Owens is a tech and media journalist living in Washington, DC. Follow him on Twitter, Facebook, or LinkedIn. Email him at simonowens@gmail.com. For a full bio, go here.
Related articles:
