Can Fb Work Out Find out how to Gradual the Momentum of TikTok?
In all fairness, after years of running the tables on social networks and improving many of the best innovations in other apps, Facebook is refreshing to see them on their backfoot and work on keeping up with the competition for a change.
I’m talking about TikTok, in this case, and The Social Network’s efforts to crack the short video app’s code to curb its growth momentum. So far, Facebook has had no success – TikTok is well on its way to becoming the next billion-user platform in 2021, and its users reportedly spend significantly more time on the app than people on Facebook or Instagram.
This is an unusual position for Facebook. The last time it competed against this level of competition, Snapchat gained momentum with younger viewers, which at one point posed a long-term existential threat to Facebook’s dominance. Instagram Stories set and finish that – but according to the same playbook, this time with Instagram Reels, it doesn’t seem to have worked yet. And Facebook is still trying to figure out why exactly that might be so.
The main problem, however, is probably something that Facebook and Instagram cannot overcome. The construction of TikTok and the way users are interested in it is fundamentally different from Facebook’s approach to interest graphs.
Facebook has so far been able to defeat opponents due to its ubiquity and huge data repositories that allow people to show more of what they like, of the people they care about most. Its algorithm is convincing because it knows who and what you are engaged in – but TikTok, which, unlike closed networks, focuses on public publishing, opens its content pool to the millions of videos that are uploaded to its app every day, whatever it enables him to identify trends and interests more precisely and target the feed to many more clips in that direction.
And it’s not just the breadth of the content – TikTok only shows users one full-screen video at a time, which it can use to more closely track certain signals of interest, which in turn improves the recommendation engine.
As industry analyst Eugene Wei noted:
“Anything you do from the moment the video starts playing is a signal of your attitudes towards that video. Are you swiping to the next video before it even plays? An implicit (if marginally explicit) signal of disinterest. Have you seen it more than once and repeated it a couple of times? Seems you liked something about it. Did you share the video using the built-in sharing area? Another strong indicator of a positive mood. “
TikTok’s algorithm, which includes machine learning based on video content as well as human categorization, is highly tailored to people’s specific interests and requires little effort to learn what is addicting. Because of this, you will scroll into the wee hours of the morning and feel fresher than your regular social feeds.
Facebook and Instagram weren’t constructed in the same way, although Instagram tries to replicate the approach using reels. And another element in that direction has recently been added.
As Facebook explained late last week, a new process is being developed to categorize videos based on content – similar to TikTok.
According to Facebook:
“Popular videos often consist of the same music, tuned to the same dance moves, but created and played by different people. Self-monitored models automatically learn “topics”, group them and implicitly make them available to the recommendation system. We use self-monitoring to suggest videos that are relevant to recently viewed videos and to filter out nearly duplicate data – without explicit training labels for each classification task. “
In addition, as the app researcher Alessandro Paluzzi found out, Instagram would like to add separate tabs “Discover” and “Follow” in Reels, similar to the TikTok feeds “Follow” and “For you”.
This would similarly expand Reels’ content pool while Facebook is also trying to capitalize on its optimal reach lure to instill more creators by allowing people to share their roles on Facebook as well.
Will this make Reels a stronger competitor for TikTok and stop TikTok’s general momentum?
Probably not. Here, too, the construction of the TikTok app is specifically geared towards this extremely appealing main feed. With Reels as an add-on within Insta, this is not the same, although it can improve its recommendations and tailor them more to individual interests.
Essentially, TikTok sees success by beating Facebook in its own game and using implicit signals to optimize the user experience. Facebook needs to catch up now, which is interesting to see as it ponders how to slow down the dynamics of the app and avoid losing ground with younger users.
Perhaps Facebook will instead shift its focus to AR and VR, acknowledge that shift, and try to bounce back with the next level. This seems increasingly like a more viable option as Reels is still not arriving in any way and Instagram is feeling more and more crowded with various features.
That probably means TikTok is staying here. And it could become a more important consideration for marketers moving forward.