Kayla Dunn,
Potomac Institute for Policy Studies Intern, Summer 2024, and Student, Georgetown University
It is now well understood that social media platforms collect data about their users. The platforms gather this data to push content that supports targeted advertisements. This business model incentivizes platforms to maximize user engagement and disseminate content that negatively impacts society. Notably, misinformation, disinformation, and mal-information demonstrably generate higher levels of engagement.
Platforms claim they collect user data to personalize and improve the user experience. For example, TikTok and Meta platforms utilize data to craft their “For You Page” and “Recommended Posts,” respectively. However, this language obscures other motivations for user data collection, including targeted advertising. Today, no one doubts that these platforms collect user data, but the extent and purpose of this collection are often overlooked.
Social media platforms collect user data in four ways:
User-provided data: Data a user produces, including posts, searches, comments, “likes,” content viewed, content engagement levels, and user profile information such as email addresses, phone numbers, and contacts.
Device information: Data from the devices users use to access the social media platform, such as device model, software, location, photo access, and network connection information.
Data collected from other users: Posts by other users on the platform that tag or mention the user, networks of users including “friends” and who is “followed,” and search histories that involve users.
Third-party data: Data shared from external sources, such as other platforms and browsers, websites visited via social media platforms, cookies, and third-party location.
Ad targeting begins when an advertising company contacts a social media company with an ad campaign. The advertising company has a specific demographic they want to target. The social media company then asks the advertising company about their budget for the advertising campaign. In essence, the advertising company bids for advertising space for their campaign, with a separate price for a liked post, an ad that is followed with a click, a website that is visited, and potentially even a sale that occurs as the result of the ad on the social media company’s platform. The social media company can execute the auction as real-time bidding, or have predetermined prices in advance for targeted advertising.
Either way, the social media company is now highly motivated to (1) have as many people as possible of the targeted demographic spend plenty of time on their platform, and (2) to present the ad to precisely targeted users to maximize what is called “click-through rate,” i.e., the likelihood that the ad will generate revenue through a charge to the advertiser.
Advertising rates increase proportionally to users’ engagement level on the platform. When the social media company optimizes its targeted demographics, clicks will be more likely and more frequent. To achieve high levels of engagement, social media companies benefit when a post or video goes “viral.” Every interaction can refine algorithms targeting users. This transactional relationship places utmost importance on social media platforms increasing engagement levels through continuous manipulation and revision of platform targeting approaches. This process is meticulous. It encourages viral posts and sustained interactions.
Further, the quality of content is not the primary concern of social media companies; their main interest lies in the level of engagement with their content and the likelihood of ads generating revenue through “clicks.” Research confirms that content achieving the highest levels of engagement is mis-, dis-, and mal-information. On Twitter (now known as X), studies have shown that “falsehoods were 70% more likely to be retweeted than the truth.”
On TikTok, a 2022 study by NewsGuard found that “almost 20 percent of the videos presented as search results contained misinformation.” On Facebook, a study conducted by a joint research team at the University of Southern California found that “frequent, habitual users forwarded six times more fake news than occasional or new users.”
This is a symptom of the social media business model; content that garners engagement is rewarded regardless of quality.
Social media companies and those selling ad space typically utilize proprietary AI algorithms that leverage collected data to profile and characterize each user. This has revolutionized the advertising business. “Data commercialization” includes the entire process of developing and presenting content, and collecting, characterizing, and placing targeted ads.
The effect of data commercialization is evident through the rapid growth of social media advertising revenue. In the two decades following the founding of Facebook in 2004, social media advertising has become a multi-billion-dollar industry. For example, Meta earned $38.7 billion in advertising revenue in Q4 2023 and Tik Tok generated $14.5 billion in advertising revenue in 2023. Another estimate suggests TikTok generated $16.1 billion in 2023, a 67% increase year-on-year. Currently, global spending on social media advertising is approximately $270 billion, with that figure forecasted to increase to $345.73 billion by 2029. Advertising in the United States has consistently remained around 2% of GDP since the 1920s. Social media companies and online platforms are taking an increasing share.