When algorithms triumph, who suffers a loss?

In a post-social era, recommendation media offers consumers a better consuming experience, but individual outcomes may differ. I wrote The End of Social Media last week, explaining the reasons for platforms' move away from social graphs and toward algorithmic, recommendation-based methods of content delivery. Here is the TLDR if you haven't read it yet:

When algorithms triumph, who suffers a loss?

For platforms, sharing material through friend graphs is ineffective. More significantly, it creates an opportunity for adversaries to develop more effective models and generates enormous expenses in the form of sizable moderation teams and serious brand harm to platforms.

Because platforms, not authors, choose what is viewed and when, recommendation media, which disseminate material through user-targeted algorithms, are more effective, more defendable, and less vulnerable to misuse.

Platforms will likely utilize synthetic media to provide the ideal material for each user at the ideal moment as they seek even more control and efficiency in feeds.

The landscape of media platforms is enormous, and several parties are involved in the production, distribution, and consumption of material online. It is obvious that the platforms themselves are the clear winners in a world where recommendations replace friend graphs and profit from the paradigm change.


However, who loses? Which stakeholders' operations are most likely to be affected by this significant change in content distribution? Let's begin...



One of the primary media types on social media for a very long time has been images. A person could easily share images of family trips or photos with their friends from last night's party in the early days of social media, when pictures arrived in the form of family photos thrown into enormous Facebook albums. Then, Instagram turned everyone into an artist by providing stunning filters that allowed quick and easy picture editing. However, Snapchat was responsible for making images a genuine form of social media communication with vanishing photographs and, eventually, stories (both boosted by an arsenal of fun tools to reduce the friction of sharing photos). As a consequence, exchanging images has become as common and informal as texting.


It's no secret, however, that videos have consistently outperformed other types of media in terms of interaction, with almost every platform increasing its emphasis on the format in recent years. Videos, by their very nature, transmit far more context and information, and, as a result, require greater attention from viewers. We should all anticipate seeing a lot more video in our feeds if recommendation media is only about content distribution with the aim of maximum interaction. This will unavoidably come at the price of images and influencers who rely on sharing images on social media sites like Instagram to grow their fan bases (and professions). As a consequence, I anticipate that if they have trouble finding a market for their images, many photographers may go into alternative avenues for artistic expression.



However, everyone who has spent a lot of effort increasing their follower count will be affected, not just the influencers who share photos. It's understandable that Kylie Jenner, one of the most popular social media users, would be opposed to the move to recommendation media, as I briefly discussed in my earlier article, since she would have a lot less programming authority. Less programming power leads to lower content engagement, which reduces the desire of sponsors, companies, and advertisers for access to her audience.


However, recommendation media won't simply have an impact on Kylie Jenner and the largest influencers in the world; it will also lessen the total worth of a single follower across all significant platforms. Millions of individuals who have invested years in building an audience in order to distribute material will need to reconsider (or give up) their methods in favor of new playbooks that put the development of successful content ahead of personal brand loyalty. Given the secrecy around the precise factors driving interaction through platform algorithms, this will be particularly difficult for artists. The social media blueprint was straightforward: develop a following, get distribution. Instead, the strategy for recommendation media will be to provide content and cross your fingers. This perspective makes it simple to understand how the influencer marketing industry is poised to undergo significant transformation.


family and friends.

Let's not lose sight of the main objective of social media, which was for many of us to communicate with our loved ones. Social media has played a significant part in our capacity to remain connected as a species over the last several decades, despite the drawbacks of friend-graph-based content distribution (such as "guaranteed circulation" for problematic information, echo chambers, etc.). And as more of us migrated to spending our lives online over time, notably during the COVID-19 epidemic, the demand for this kind of distant connectivity only grew.


It might not be as easy to share life updates with one another for much longer. In the days to come, we will be able to confidently upload a picture to Instagram and share it with many of our friends and family. In contrast, a valuable video from a total stranger may push the identical picture out of the stream in recommendation media. Because of this, I anticipate that individuals will be much more deliberate about when and where they share personal material with friends, preferring private messaging services like iMessage, WhatsApp, or Messenger instead of reviewing websites.



A large, broad library of material and top-notch machine learning algorithms are, generally speaking, the two essential components of a successful recommendation engine. While the latter is required to carry out intelligent matching amongst constituents, the former is necessary to guarantee that each unique consumer on the platform may be precisely matched with the material that best meets their individual interests. Both need a huge size and amount of platform capital, which the main platforms currently have. However, in a world of recommendation media, entrepreneurs attempting to compete with the platforms would face a larger disadvantage. In contrast to many new social networks, which depend on friend graphs to spread material, the platforms will be able to perfectly match content and users with a great deal more efficiency because of their top-notch ML.


On the other hand, this new dynamic might also make it easier for firms focused just on social media to become relevant. While it's obvious that the main platforms think that algorithmic content distribution would improve their company, it doesn't necessarily indicate that a challenger can't have a successful business model using social distribution. New firms will try to fill the gap in human connection that can grow if our newsfeeds include less material from our friends and family. With purely social applications like BeReal topping the App Store rankings, we may already see this happening to some degree. To avoid being readily imitated by the mainstream platforms, these new platforms will need to accomplish something really original with their format in order to remain relevant.

What's Your Reaction?