Arthur Musgrove Home

Engagement and Usage Decline in Facebook

Synopsis: Time spent on Facebook has been in decline since well before their data privacy crisis began. This has implications for developers, publishers and advertisers. This article analyses this impacts of the Facebook changes announced 11th January 2018.

Updated: 19/April/2018

The Situation

When Facebook announced it's algorithm changes on 11th January 2018, they predicted a 5% reduction in user time, or 50 million hours per day (yes, day!) in user engagement time. While good data has been hard to come by, it appears the actual impact has been far worse. Neilsen Digital Content ratings reports a massive 24% decline in average time spent on the platform. In the US, the number of daily active users has reduced from 185 to 184 million (source). That may seem like a small decline, but it is the first decline.

The declines began before the Facebook data privacy crisis. In the aftermath of the "Fake News" crisis which was highlighted most vividly in the dissemination of misinformation during the UK Brexit and US Presidentatial campaigns in 2016, Facebook came under enormouse pressure to change it's Newsfeed algorithm - the bit of software that determines what you see and how prominent it is. This may sound like a bit of geek stuff, but it is having a profound effect on Facebook, the media industry, Internet and business.

An Engagement Driven Business

98.5% of Facebook revenue is from advertising (source). As measured by Statistica, a full 18% of global ad spend goes to Facebook. Advertising is core to Facebook, and Facebook is a key part of any online advertising strategy.

Against this background, Facebook's business model has always been to increase the amount of time spent on it's platform. That makes it more valuable to advertisers. It does this by tuning your newsfeed so you see things likely to keep you on it's platform. It's share price has always been roughly a function of the number of hours users spend staring at the newsfeed - and so being exposed to ads.

Facebook delivers value to it's advertisers by showing users content that will keep them on the platform for more hours every day. Who doesn't like watching the latest cat-that-can-dance video? And for users that create content that people widely want to see - meaning gets them to spend more time on Facebook - Facebook will pay them through their Content Monetization program. Yes, they will pay you for making funny cat dancing videos.

So far, so good. But then malicious actors started feeding unsavoury content into the platform. The most infamous of this has been the politically charged "Fake News". That charged content, playing to people's inner biases and emotional feedback loops, did on it's surface exactly what Facebook wanted - generated rapidly increasing time users spent on the platform. It was also coorosive, so Facebook came under unbearable pressure to make changes that go to the core of it's business model.

Algorithm Changes

Facebook's answer was to change it's algorithm to give what it called more "meaningful engagement". That means showing more things in your newsfeed about your friends and family, and less commercial and public content. Public content would also be prioritized by it's engagement, generally measured by things like comments, shares, and conversations between friends occuring in the context of the content. Media content that was being shared would also start to be prioritized by the quality of the source, not merely how many 'clicks' it garners.

To the average user, these changes sounded commonsense - and surely they are - and as if the impact would be minimal. The impact has been far from minimal on either publishers or advertisers.

Impacts of the Changes

Some of the most immediate and dramtic effects have been noticed in the media business, as noted by online media analytics firm NewsWhip. With Facebook limiting the number of public content posts and prioritizing sources generally more reputable, better known and established outlets have gained and fringe content has seen a marked decline. Examples of increases include CNN, which saw a dramatic 31% increase, as well as the New York Times with an even more dramatic 51% increase in interactions. Given overall user time and engagement on Facebook has decreased in the same period, if you guess these increases have come at the expense of fringe and lower quality content, you'd be right. There are a host of previously highly trafficed but lesser known properties that have seen dramatic falls, such as Western Journalism and NPR.

On this topic of changes in traffic, there have been widespread online conspiracy theories that Facebook is discriminating against politically 'conservative' content in favor of politically 'liberal' content. I have no intention of doing political commentary, but given I've done the research on the data for this article, I'd feel remiss if I did not point out that Fox News - known to be Conservative - has seen a marked increase in engagements after the changes, where NPR - widely seen as Liberal - has seen one of the most dramatic declines in engagements (around 40%) of any major sites. It's a fun parlor game to speculate on who, what and why these major traffic changes are happening beyond what Facebook says publicly, but the data at least suggests it isn't all politics.

Beyond media, advertisers have felt the pinch. As Facebook has prioritized user generated content from our connected friends and family, less 'real estate' is available on the newsfeed for ads. Those effects are already being noticed and are generating a lot of industry press. I won't repeat what is already written, but you can see an interesting article from AdWeek on the topic back in January. From a commercial standpoint, user's now see less ads, which may make the advertising space on the newsfeed more valuable. However, there is also less overall user engagement, tending to make advertising space less valuable. How will it balance out? It is wait and see for now.

And remember those people making money from cat videos? The final major area of adverse impact, which has been far less reported on in the media, are the people that earn income by monetizing content. There was a story of online commentators known as Diamond & Silk. After the algorithm changes they suffered a decline and traffic which very negatively effected their income from monetization of content. While they started a hoax conspiracy theory that they were being discriminated against that in the end was debunked, their underlying observation was true: the algorithm changes did indeed reduce their traffic, engagement numbers and consequently ability to monetize their content. They are far from the only ones.

Reactions In the Market

The effects on online businesses, advertisers, publishers, and content providers of the changes in Facebook's algorithm are only now starting to be aparent, and obviously Facebook will make adjustments. A sure beneficiary are digital marketing firms as they help clients adjust. Other beneficiaries may be other platforms able to more micro-target ads from certain supplier or provider segments into very specific customer segments. Facebook had a stumble in it's attempts at micro-targeting users with the Cambridge Analytica scandal, but that in no way means the increasing ability to micro-target ads and content will become unattractive in the online industry.