UEIt's literally the billion-dollar question: How did TikTok go from a niche social network for lip-syncing teens to the most popular app in the western world that's threatening to collapse?Facebookcompletely out of the question in a few years?
There is no end to the possible answers, andtag thank youowes its phenomenal success to a variety of smart decisions: easy-to-use video creation tools have blurred the line between creator and consumer much more than YouTube; A huge library of licensed music allowed teens to add music to their clips without fear of copyright. A billion-dollar ad campaign on Facebook and Instagram brought in new users as fast as Zuckerberg's company sent them out.
But the most powerful tool TikTok has to engage users and keep them coming back is the company's acclaimed "Page for You," the FYP, and the algorithm that completes it.
FYP is the default screen new users see when they open the app. Even if you don't follow any other accounts, you'll find yourself instantly filled with an endless stream of curated short clips of what's popular across the service. This decision has already given the company an advantage over the competition: a Facebook orGoreAn account with no friends or followers is a lonely and barren place, but TikTok has been captivating since day one.
However, what happens next is the company's secret ingredient. As you progress through FYP, the composition of videos presented to you will slowly start to change until, as regular users of the app say, it becomes almost unbelievably good at predicting which videos on the site will pique your interest.
The company is surprisingly candid about how this algorithm works, at least on the surface. "The recommendations are based on a number of factors,"these 2020, “including things like user interactions like videos you like or share, accounts you follow, comments you post, and content you create; video information, which may include details such as subtitles, sounds, and hashtags; [and] device and account settings, such as preferred language, country settings, and device type.”
But how these different inputs are weighted and what exact factors lead to a given video ending up in your feed is opaque, says Chris Stokel-Walker, author of TikTok Boom. "A person at TikTok responsible for tracking what's going viral and why told me in my book, 'There's no recipe for this, there's no magic formula.' The Algo team has the answer to that. It's so sophisticated".
One important innovation is that, unlike older recommendation algorithms, TikTok doesn't just wait for the user to give a thumbs up to indicate that they like a video or that they are satisfied by judging what the user is watching. Instead, it seems to be actively testing its own predictions, experimenting by showing videos it finds entertaining and gauging the reaction. "He pushes the boundaries of his interests and monitors how he interacts with the new videos he puts on his For You page," Stokel-Walker says. "If he thinks you like Formula 1 videos, he might show you some supercar videos."
This experimentation not only allows the service to quickly discern the contours of an individual viewer's interests, but is also an important part of what the site offers creators, says Sascha Morgan-Evans, director of TikTok studios at the OK COOL creative agency. “Every video posted on TikTok will be seen by at least one person on the For You page. Based on the number of views, we found that TikTok offers all the videos to many people. The number of users in those batches increases with each successful round where the majority of users in a batch had a high number of positive interactions with the video.”
This means that every user has a chance to achieve world fame. Even if you don't have a following, your video will eventually land on someone's For You page and, assuming positive engagement, you can reach thousands or millions of viewers extremely quickly. And the speed of videos helps TikTok refine its data quickly: "Think about how many videos you watch on YouTube in an hour," says Stokel-Walker, "and what data is generated about you, compared to how many you can watch on YouTube." ". TikTok”.
However, FYP is not magic, and the ways in which it fails can be just as revealing as the ways in which it succeeds. New users of the app will discover that it is obsessed with collecting personal information, requesting contact list access, and tracking every video shared. Denying him these data points forces him to present the most generic version of the feed possible, customized with what little he can determine from general geolocation and device details.
But when it works, the algorithm is so good at what it does that TikTok seems almost overwhelmed by its power, Stokel-Walker says. “He even messaged users he considers very addicted and told them to turn off.”
One such message, displayed by the company's TikTokTips account to users who scroll through their feed for hours at night, shows TikTok star Gabe Erwin pleading with the viewer: "Go get some more sleep, turn off your phone". please and have a great night.” The company has also added new “Screen Time” features, especially for younger users.disable notificationsextend past bedtime and allow users to do soset max timein the app every day to limit more compulsive use of the app.
As TikTok enters its second year of online dominance — the app surpassed YouTube in average time per user in September 2021 and has stayed on top ever since — the big question is whether its algorithmic success can continue to be a differentiator. Facebook certainly isn't expecting this: the social network, along with its sibling Instagram, recently announced an overhaul of its apps to focus on an aggressive new algorithmic curation engine. Like TikTok, Facebook and Instagram now display massive amounts of content from users you don't follow, with friends' posts interspersed or hidden behind a tab in a separate "Following" feed.
the change wasannoying, leading to an apology video from Instagram boss Adam Mosseri, who said the company would remember some of the changes, but that's the future. "We're going to try to improve recommendations," he said, "because we think it's one of the best ways to help creators reach new audiences and grow their following."
If there is a threat to TikTok's algorithmic crown, it could come from the company itself. The app dominates user attention, but has seen little monetization in the past. As a private company, TikTok doesn't release revenue figures, but as of 2021, research firm eMarketer estimates it earned $4 billion a year, less than 5% of Facebook's revenue.
In 2022, the company tried to increase this. It took the traditional approach, putting more ads in the feed, but also tried new ways, including a push for QVC-style live shopping experiences carried over from Chinese sister app Douyin. The beginning was bad. Hosts and brands had to be subsidized by TikTok, which resulted in heavily discounted sales, but failed to attract a repeat audience.
Worse still, sales at too-good-to-be-true prices were hurt by other too-good-to-be-true priced items: the platform wrestled with a counterfeiting problem, leaving users wondering if a Dyson hair dryer was worth £ 1,450 sells for £14 due to subsidy or fraud.
But if TikTok can figure out how to balance commercially necessary adjustments to its algorithm with sheer FYP coercion at its best (or worst), then it has created an artifact of technological history that will perish along with News Feed, Infinite Scroll, and Snapchat. ) History as milestones of the era of social networks.