I’ve been a curator of online community spaces for nearly a decade now, which means I spend more time than most people staring at the system of what gets seen and what doesn’t. I watch which posts gain traction and which ones disappear into nothing, and, historically, the inflammatory one or hottest take almost always wins the social media game because engagement and virality is the currency these platforms are built on.
The business logic is not complicated here — make high engagement content more visible, show the metrics, sell more ads. In 2021, whistleblower Frances Haugen confirmed through internal documents that Facebook’s own research showed that the company had weighted angry emoji reactions as five times more valuable than a standard like when determining what to promote in your feed.
Facebook’s own data scientists flagged the problem almost immediately: Posts that generated anger were disproportionately likely to contain misinformation and low-quality content. The company made adjustments over the following years, but not before that weighting had reshaped how billions of people saw information across the platform. Haugen told the British Parliament directly that “anger and hate is the easiest way to grow on Facebook.” A landmark MIT study published in Science tracked 126,000 news stories across Twitter over more than a decade and found that false information is 70% more likely to be shared than true information, and reaches its first 1,500 people six times faster. It wasn’t bots driving the behavior, it was humans.
In 2022, researchers at Wharton found that ad-supported platforms have a structural reason to moderate less aggressively than the public interest requires because engagement and safety online run contrary to each other on social media. Every post removed is a potential interaction that never happened and an impression never to be sold.
When performed with fidelity, content moderation is expensive and revenue-negative, so I don’t think anyone was surprised when in early 2025 Meta announced it was ending its third-party fact-checking program, replacing it with a crowd-sourced community notes system, and rolling back content restrictions on political topics. Zuckerberg was candid about what the change meant: “We’re going to catch less bad stuff.” By midyear the company posted a 22% revenue increase.
Perhaps the more hyperlocal version you may be aware of is Nextdoor, which presents itself as a neighborhood resource but still drives revenue through ads and engagement and uses community volunteers to moderate their respective corners of the internet. A peer-reviewed study published this spring in the journal New Media & Society by University of Colorado Boulder researchers Toby Hopp and Patrick Ferrucci reports that regular Nextdoor users report significantly higher concern about crime and show greater openness to aggressive policing tactics than people who use the platform less or not at all, even as national crime data shows overall declines.
Recommended for you
Different platform, same revenue model, same engagement triggers, same human outcomes. Nextdoor’s own CEO once said in an interview that more neighbors using the platform means more ad inventory and more revenue for the company.
Where so many people go to get a sense of whether your neighborhood is safe and whether your neighbors are trustworthy is being engineered by people whose incentive is not truth but engagement. Every conversation is filtered through algorithms optimized for engagement, data and revenue.
California’s June primary just wrapped, which means you’ve spent the past few months absorbing increasingly intense social content designed to trigger emotions across races up and down the ballot because that’s what still wins in the “pic for algo” game. While we are in this temporary lull of all caps posts with AI generated imagery and a sprinkle of unverified accusations intended to fuel online outrage, now is the time to remind ourselves that what registers as community conversation on social media is also always an ad product.
Today, misinformation is dressed up to look like the truth, created with AI tools that make it easier than ever to paint someone as a caricature, appropriating language that you’ve heard before in another context, and used by people you trust. The only antidote to yet another train wreck of social media politicking is human presence and putting in the work to pick up the phone, send an email or grab a coffee and have a real conversation.
Our unfortunate reality is that the version of people and community you’re seeing online has been curated by people whose job is not to inform you but to keep you engaged long enough to get more data and sell an ad. Yes, scrolling and tapping on a screen from bed is a lot easier than reaching out. But this is the work that’s going to preserve your relationship to the people around you and your ability to make a clear decision when you vote in November and beyond.
Annie Tsai is chief operating officer at Interact and three-time author, leads community engagement and learning for Moms in Tech, and is a city and county commissioner, among other things. She can be reached at: media@annietsai.co.
Keep the discussion civilized. Absolutely NO
personal attacks or insults directed toward writers, nor others who
make comments. Keep it clean. Please avoid obscene, vulgar, lewd,
racist or sexually-oriented language. Don't threaten. Threats of harming another
person will not be tolerated. Be truthful. Don't knowingly lie about anyone
or anything. Be proactive. Use the 'Report' link on
each comment to let us know of abusive posts. PLEASE TURN OFF YOUR CAPS LOCK. Anyone violating these rules will be issued a
warning. After the warning, comment privileges can be
revoked.
Please purchase a Premium Subscription to continue reading.
To continue, please log in, or sign up for a new account.
We offer one free story view per month. If you register for an account, you will get two additional story views. After those three total views, we ask that you support us with a subscription.
A subscription to our digital content is so much more than just access to our valuable content. It means you’re helping to support a local community institution that has, from its very start, supported the betterment of our society. Thank you very much!
(0) comments
Welcome to the discussion.
Log In
Keep the discussion civilized. Absolutely NO personal attacks or insults directed toward writers, nor others who make comments.
Keep it clean. Please avoid obscene, vulgar, lewd, racist or sexually-oriented language.
Don't threaten. Threats of harming another person will not be tolerated.
Be truthful. Don't knowingly lie about anyone or anything.
Be proactive. Use the 'Report' link on each comment to let us know of abusive posts.
PLEASE TURN OFF YOUR CAPS LOCK.
Anyone violating these rules will be issued a warning. After the warning, comment privileges can be revoked.