Close
Close

Published

Facebook closes video gap on YouTube, has greater ambitions

Facebook overhauled its video strategy during the second half of 2017, but it was already working quite well if we can believe various surveys that show it has been closing the gap on YouTube for mobile views. The latest survey, admittedly without any indication over methodology, from Openwave Mobility, a provider of traffic management software for MNOs, suggests Facebook is posed to overtake YouTube worldwide for mobile views and has already done so in many developing countries.

This highlights the success of Facebook’s mobile app and also the first stage of its video strategy which began in earnest in 2014, focused on engagement around very short form video mostly published in its News feed. The aim was both to increase time spent on the site at the expense of others and to generate revenue from mid roll advertising.

The content was at first a motley collection of User Generated Content (UGC) and semi-professional material, but then more recently Facebook encouraged fully professional outfits to post content for user’s news feeds alongside personal material, which was not a complete success as it coincided with the 2016 US presidential election campaign. It led to Facebook being the forum for posting a lot of “fake news” from Russian sources or otherwise, which led to condemnation and negative publicity.

At the same time Facebook chief Mark Zuckerberg realized that this first phase video strategy, despite being by some measures highly successful, had gone as far as it could and needed a complete overhaul to start reaching towards his ultimate goal. That is to go beyond competing with YouTube and other social medias sites, or even with pay TV operators, to become the place of TV itself, the virtual living room of the post OTT era. If that sounds like hubris it is why Zuckerberg has tried to restrain expression of this ambition, but that’s what it is.

At the same time Facebook needed to counter the issue of fake news and answer the criticism that its video strategy was too fragmented and small time. However there is still a tension between the obvious utility and revenue associated with the ultra-short form video that is almost its legacy and the longer form, higher quality video it is now seeking to attract and indeed produce itself.

The first step came in February 2017 when it started funding video from well known partners like BuzzFeed and the New York Times, posted on its news feed according to analysis of user preferences. This came with Zuckerberg’s announcement that Facebook was pursuing a “video first” strategy across all platforms focusing strongly on mobile. This strategy would fork in two directions, hence the slight confusion, with on the one hand an expansion of short form production and an increasing focus on ads. On the other hand the budget available for video producers to create longer, higher quality scripted and unscripted projects was increased. Facebook has set its original content budget at $1 billion for 2018, matching Apple and a big rise, overcoming early reluctance to spend heavily on acquiring programming.

Announcement of the 2018 budget coincided in September 2017 with launch of the new Watch video platform, positioned more directly against YouTube as a place where users can browse and seek out content rather than just have it recommended to them. At the same time the News feed was reorganized, which is about to come into force after an announcement early this year. This aspect of the revamp, splitting out videos into two platforms or tabs within the site, is partly to show Facebook being the good citizen and responding to the issue of fake news.

Facebook has conducted some research into how users would like the whole service to evolve. The main finding, hardly a great surprise, was that users would like the service to connect them more actively to people and perhaps even issues and services as well that they really care about, rather than having items that they have not solicited thrust at them. This has led to redesign of the whole site to favor more active engagement rather than passive actions, such as “likes” and “shares”. As Zuckerberg put it recently, the new emphasis is on helping users have “meaningful actions” rather than finding “relevant content”.

Quite what this means is not clear, but Facebook has tweaked the Watch algorithm and made bigger alterations to the News feed algorithm, although with the indication that further modifications will be made in the light of feedback after the first changes go live. The main difference for video is that the News feed will now comprise items recommended by closer friends, which could still include professional content, but with supposedly less chance of fake news appearing there, helping to appease critics. Then a lot of the videos that used to appear in the News Feed will now be under Watch. That will also be where the longer form higher quality content will appear, unless recommended or shared by closer friends for posting to the News feed.

The algorithms are also being revamped to change how videos are prioritized. There will be greater emphasis on repeat viewing, so that videos with high view counts or that are part of series will be shown more prominently in users’ News Feeds. The obvious aim is to encourage publishers towards more episodic content that builds engagement.

Then combining the two platforms, Watch and News Feed, Facebook is prioritizing videos that users search for proactively. The idea is that videos heavily searched for on Watch will be more widely distributed on users’ news feeds. The ultimate aim is to persuade users to seek out Watch content actively rather than passively viewing short-form videos in the News Feed. So this is part of the strategy to steer users towards more active engagement, partly because according to its research it makes them happier, but also because it encourages them to spend more time on the site with deeper engagement.

We have not heard much about Facebook’s R&D into content search which must be just as relevant as it is for Google and YouTube. It is likely it will be able to leverage investment being made into techniques for recognizing fake news and other undesirable content of terrorist origin, which also requires content aware searching techniques.

The other significant video-related changes Facebook has made recently are on the advertising side, especially over placing and timing of mid roll ads. This is part of the strategy to increase viewing time and attract more ads accordingly. Until now mid-roll ads have played after 20 seconds on videos at least 90 seconds long. This is being changed to playing them after one minute on videos at least three minutes long, with Facebook arguing this will promote stronger engagement with the ad. But some advertisers have complained that users will not tune in for as long as a minute to view mid-roll ads and so revenues or engagement are reduced rather than increased. The average time users spend watching a Facebook News Feed video is after all only 17 seconds at the latest count.

More noticeably for users Facebook is starting to test pre-roll ads, but only against videos on the Watch platform and not the News feed. Facebook has shunned pre-roll ads so far on the grounds that users do not like them and find them disruptive on a social media platform where the videos are very short anyway. But it makes sense at least to test them on Watch, given Zuckerberg’s ambition of making this platform self-financing. For that reason it is unlikely that the budget for original content will match the levels of Netflix or Amazon over the next two or three years, while the aim is also to ensure that third party content does not infringe stricter controls over quality and authenticity.

Against this background it makes little sense comparing Facebook with say You Tube via any metric other than revenues, considering we are still comparing apples with oranges. Video consumption can be measured by traffic generated, viewing of the first three seconds, rating, average watch time, length per view, or completion rate, but none make sense for all types of video. One undeniable statistic is that Facebook has the best collection of data given its base of 2.2 billion active monthly users with intimate knowledge of preferences and friendship groups. That has already made the company a strong force and combined with the increased investment in content that data could make it irresistible.

Close