Social content based on relationships, or social content recommended based on interests, which one do you prefer at present?

At the end of July this year, the initiative of “Make Instagram Instagram again (let IG be IG back)” aroused heated discussions on the Internet.Actress Kylie Jenner, who has the most followers on Instagram, also posted a comment expressing her strong dissatisfaction with Instagram’s feed that prioritizes video recommendations over friends’ photos.

As the “little golden sister” of the Kardashian family, Kylie Jenner has more than 360 million followers on ins, and her influence cannot be ignored. Faced with more and more questions from users, Instagram CEO Adam Mosseri later released a video discussing some recent ins innovations and future plans. In the video, Mosseri says the world is changing and Instagram has to change with it.

In the world of social media, one of the changes Mosseri said is exactly what Kylie Jenner complained about – the content distribution logic of social media is changing from recommendation based on social graph to algorithm based interest recommendation.

At the end of July, Meta, the parent company of Instagram, officially announced that the information flow of another social giant, Facebook, will shift to a content distribution model based on algorithmic recommendation. Some researchers believe that Facebook’s official pivot heralds the end of “social media” as we used to know it.

This article will take you to know behind Facebook’s turn to algorithmic recommendation to social media, what we call “social media”, what two decades have passed, and where is it going?

social media

From Facebook to Twitter, an “empire” built on the social graph

Last year, Facebook added Reel, a TikTok-like short-video section to its app; in an internal memo that leaked this spring, Tom Alison, the social media giant’s senior management, announced new plans to change Facebook’s the content distribution model makes it more focused on short videos, and the algorithm is adjusted to push the most engaging content, even if the content has nothing to do with the user’s friends or followers being pushed.

It appears that Facebook is changing its focus, from graphic attention around the social graph, to borrowing from TikTok, skewing traffic to content that algorithms calculate,which are the most to grab users’ attention.

Facebook’s shift isn’t surprising given the phenomenal popularity of TikTok.

To understand where Facebook is now, perhaps we should first look back at how Facebook was successful nearly two decades ago.

When college students at the time signed up for The in the spring of 2004, it was mainly because other people they knew were also signing up for the site.

One of Facebook’s early killers was the ability to view a classmate’s “relationship status.” By the end of 2006, the year Facebook opened up to the public, the site had amassed 12 million active users.At that time, the growth advantage created by the Matthew effect made it difficult for competitors to challenge Facebook. Two years later, when Facebook surpassed 100 million active users, that competitive advantage continued to expand – if everyone you knew was already on Facebook, would you bother to join a new platform to reconnect acquaintances?

After Facebook, the second phenomenon to use the social graph to create engagement is Twitter.Although Twitter was born in 2006, it didn’t gain widespread attention until 2009. This year, American actor Ashton Kutcher discussed Twitter on The Oprah Winfrey Show.

It was also in 2009 that something crucial to Twitter happened: the birth of the retweet button. This feature was originally intended to simplify the manual copy-pasting of interesting tweets, but it ended up revolutionizing Twitter.By simplifying the steps to retweet a tweet, the retweet button kicks off a frantic virality where a tweet can be pushed to a large audience in a short period of time, and its audience can grow exponentially through the topology of Twitter’s social graph.

At the time, retweets proved to be a very effective method of content distribution, pushing out the most engaging content on the platform at any given moment. This potential for mass distribution of content in a short period of time is also starting to attract more influencers to Twitter, further increasing the value of its content.

Like Facebook, the larger Twitter’s social graph, the more attractive its platform is. In 2011, Twitter followed in the footsteps of Facebook, surpassing the milestone of 100 million users.

Of course, Facebook also noticed the rapid rise of this new competitor and was quick to adjust. Between 2009 and 2011, Facebook increasingly shifted its content distribution logic from chronological to emphasizing popular content; in 2012, Facebook also added a forward-style share button to its own app, enabling third-party Content on the platform can form a Twitter-style exponential spread.

Facebook and Twitter, the most successful social media of this century, are both built on the same model of leveraging large, difficult-to-replicate social graphs to generate a never-ending stream of user-engaging content. This also defines what we used to know as “social media” – social media refers to content (text, photos, video, audio, etc.) that is disseminated primarily through social networks of connected people.

This means that, based on the creator’s social network (friends or fans), everyone’s creations are spread to some extent, and everyone can have a certain audience.

So social media is really a competition based on popularity, not necessarily the quality of the content. It favors creators who have the most friends/followers; the more followers, the greater the potential for spread and influence.

With this communication momentum, social media platforms are able to expand extremely quickly. If a platform can build a large enough social graph, like Facebook and Twitter, it already has an automated content distribution system that delivers engaging, highly relevant content to large numbers of users.

In the face of new competition, the model has proven to be strong, unshakable, and staggeringly lucrative – by June 2022, Facebook parent Meta had a market cap of $562 billion, the seventh largest in the world,which is the Biggest Most Valuable Company.

And Twitter, a relatively small social media platform, was also worth a whopping $44 billion to Musk (before he changed his mind).

However, this monopoly built on the massive social graph cannot last forever. For these social media, the good days of the past decade seem to have sounded alarm bells when TikTok suddenly rose.

The new era of TikTok, the birth of “recommended media”

As a social media, TikTok’s success overseas may be precisely because it bypasses the social moat that Twitter and Facebook rely on to survive and maintain the throne. By separating attention from social connections, TikTok can compete directly for users without first painstakingly pulling user after user to build a social graph.

Unlike Twitter, TikTok doesn’t need a large number of influential users to justify the appeal of its content.Short videos capture users on a more raw level – visual novelty, clever interaction of music and action, direct emotional expression.

Unlike Facebook, TikTok doesn’t care if you have acquaintances or friends who are also using the platform. Although TikTok has some built-in acquaintance social features, they are not the main attraction of the platform.

TikTok also doesn’t rely on its users to manually share content with friends or fans, it assigns the task of content distribution to a monstrous recommendation algorithm — in 2021, a Wall Street Journal reporter conducted a survey that created more than a hundred TikTok accounts, and then found that TikTok could target the interests of those accounts with uncanny accuracy within 40 minutes.

Medium’s columnist Michael Mignano believes that the rise of TikTok has ushered in an era of “recommendation media,” and social media is a thing of the past.

In recommended media, the social graph is no longer the primary means of content distribution. Instead, the primary mechanism for content distribution is through opaque, platform-defined algorithms to maximize user attention and engagement.

And the “attention” here is often defined by the platform and tailored for users who consume specific content.For example, if the platform determines that someone likes a movie, that person will see a lot of movie-related content because it’s tailored to that person’s attention. This means that platforms can also decide what users won’t see, such as questionable content or extreme content.

Compared to social media, recommended media is not a competition based on popularity. Instead, it’s a competition based on absolute content. In that light, it’s no wonder that Kylie Jenner is against Instagram’s new changes.In a social media dominated by algorithms rather than fans,her 360 million followers depreciated.

When traditional giants learn from short videos, where will social networks go?

With Facebook looking more and more like TikTok and announcing its official switch to recommended media, a new era of social networking is looming, and it’s hard to imagine what’s coming next.

“New Yorker” columnist Cal Newport thinks this will bring positive results, because it represents the decline of these traditional social media giants in the United States; Medium columnist Michael Mignano even believes that the era of social media is over.But as we’ve seen with previous generations of the internet, platforms will always seek greater efficiencies as technology continues to advance, and from this, perhaps some interesting predictions can be made.

For example, professional media platforms may choose to switch to recommended media platforms. Given the strength of recommended media platforms such as TikTok and YouTube, and the way traditional social media platforms are following their lead, it seems likely that professional media platforms such as Netflix will follow suit.

In fact, Netflix co-CEO Reed Hastings may have even predicted this, saying that his biggest competitors are TikTok and YouTube, not other mainstream media platforms.

However, in order to be able to accurately distribute the right content to the right people, platforms need to have a sufficiently large volume of content, including extremely long tails of content that target the interests of everyone in the world.

And the only way to have so much content is to be an open authoring platform where any user can create. So, maybe in the future, long video sites will become a platform that not only accommodates professional studios, but also allows ordinary people to create.

In addition, if the essence of recommended media is that the platform has stronger control over what content users watch and what experience they get, then it is not difficult to imagine that platforms will eventually seek higher efficiency by producing their own content. We can already see professional media platforms trying to do this (e.g. Netflix producing original content, etc.).

But to do this at the scale of open creative platforms (like TikTok, Instagram), platforms cannot rely on humans. They also need to rely on machines to create, or as Matt Hartman puts it, synthetic media.

Recently, an artificial intelligence art generator from OpenAI showed the power and human nature of synthetic media to the outside world, which can generate art works from scratch, as well as edit or modify existing works.

As the cost of AI content creation content falls, and over time, tech platforms will produce more synthetic media to create more perfect content for users when conditions are ripe.

In any case, the curtain of recommended media has been opened. In the future, we will become increasingly accustomed to making fewer explicit choices (e.g., “This is what my friends posted”) and more uncertain choices (e.g., “This is what the machine pushed me”).

In the short term, we won’t notice much difference between the two, but looking back in a few years, we may find that our content consumption behavior and habits have completely changed.

Leave a Reply

Your email address will not be published. Required fields are marked *