There is a version of today's news you will never see. It exists. Someone you know is probably reading it right now. It was assembled for them, not for you, by an editor neither of you has ever met. And when researchers finally got permission to switch that editor off, what happened next was not what anyone expected.

Who decides what lands in your feed?

Not a person. Every major platform runs a ranking system that scores each post for each user, millions of times a second. It weighs what you clicked yesterday, who you message, how long you paused on a photo, what you scrolled past without stopping. Then it builds a feed for you alone.

In 2011, internet activist Eli Pariser gave this hidden process a name: the filter bubble. To show it, he asked friends to search the same word on Google at the same time. One friend searching "BP" got investment news about the oil company. Another got pages about the Deepwater Horizon oil spill. Same word, same search engine, two different realities. Neither friend knew the other's results existed.

That is the part that unsettles people. You can see what is in your bubble. You can never see what was filtered out.

How does the bubble learn who you are?

Quietly, and mostly from habit. Researchers studying algorithmic personalization (Kaluža, 2022) describe filter bubbles as habitual. They are not built from what you say you care about. They are built from what you actually do, thousands of small signals at a time. A pause becomes interest. A skip becomes indifference. Repeat it enough and the system stops showing you whole categories of news, without ever asking.

The rules also change without warning. In 2018, Facebook announced it would prioritize posts from friends and family and posts that spark conversation over content from pages. Newsrooms around the world watched their reach drop. Readers didn't change their habits. The editor changed its mind.

What happened when researchers switched the algorithm off?

During the 2020 US election, a team of independent academics worked with Meta on one of the largest experiments ever run on a social platform (Guess et al., 2023, published in Science). For three months, some Facebook and Instagram users had their ranked feed replaced with a simple chronological one. Newest posts first. No invisible editor.

The results surprised almost everyone. People on the chronological feed spent noticeably less time on the platforms. They saw more political content, and more content from sources rated untrustworthy, not less. And their political attitudes and levels of polarization did not measurably change.

Removing the editor didn't free people from a bubble. It just handed them a messier feed.

So are filter bubbles real or not?

Both answers have evidence behind them, which is why the debate never ends. Work by the Reuters Institute (Fletcher and Nielsen, 2018) found that people who come across news on social media tend to use more news sources overall than people who don't use social media at all. The algorithm, it turns out, is surprisingly generous with variety.

But generous across the whole feed is not the same as generous to every source. The editor shows you many outlets, then decides which handful make the cut on any given day. The bubble is less a sealed wall around your beliefs and more a velvet rope in front of each publisher.

What does this mean for news publishers?

It means a follower is no longer a guaranteed reader. When someone stops interacting with a news page, the editor gradually stops showing them its stories. Fewer stories mean fewer chances to interact, and the gap quietly widens. A page can have hundreds of thousands of followers and reach only a fraction of them.

The editor can't be removed, but it can be understood. It rewards conversation, consistency and content that stops the scroll. Newsrooms that publish regularly, write posts people want to respond to, and spread distribution across several platforms and their own channels are far less exposed when any one editor changes its rules.

FAQ

What is a filter bubble? A filter bubble is the personalized selection of content an algorithm shows each user, based on their past behavior. Because the selection is invisible, users can't see what has been filtered out.

Who coined the term filter bubble? Internet activist Eli Pariser introduced the term in 2011 to describe how personalization shows different people different versions of the internet.

What is the difference between a filter bubble and an echo chamber? A filter bubble is built by an algorithm. An echo chamber is shaped mainly by people's own choices of who to follow and what to read. The two often reinforce each other.

Does switching to a chronological feed escape the filter bubble? A large 2020 experiment on Facebook and Instagram found that chronological feeds reduced time spent on the platforms and increased exposure to political and untrustworthy content, but did not measurably change users' political attitudes.

Why don't all followers of a news page see its posts? Platforms rank each post for each person based on predicted interest. Followers who haven't interacted with a page recently are less likely to be shown its next post.

Summary

A filter bubble is the personalized, invisible selection of content an algorithm shows each user, a term coined by Eli Pariser in 2011. Personalization is built from habitual behavior such as clicks, pauses and skips (Kaluža, 2022). A 2020 experiment with Meta found that replacing ranked feeds with chronological ones reduced time on platform and increased exposure to political and untrustworthy content without changing polarization (Guess et al., 2023). Social media news users tend to encounter more sources overall than non-users (Fletcher and Nielsen, 2018), but individual publishers compete for a place in each user's feed. Followers who stop interacting with a news page are gradually shown fewer of its stories. Publishers reduce their exposure to algorithm changes through consistent posting, conversation-driven content and distribution across multiple platforms and owned channels.

Ready to make sure your stories make the cut? Start with Media Gridz