Facebook changes News Feed to automatically detect “clickbait” headlines, rank stories lower if they “withhold or distort information”
and there's only one way to find out which. http://www.nytimes.com/... #picks Chorus.Fm / @chorusfm : Facebook Changing Algorithm to Punish Clickbait http://chorus.fm/... http://twitter.com/... Jason Fields / @jasonqfields : They also seem to be talking about bounce rate. A lot of news stories give you what you need in graf 1, not sticky. http://twitter.com/... Matt / @mattbuchanan : eliminating clickbait seems like a pretty good reason to just show more videos in the newsfeed instead of articles http://www.theverge.com/... John Herrman / @jwherrman : every link eventually grows old and wise and realizes it was clickbait all along http://newsroom.fb.com/... Harry McCracken / @harrymccracken : If your media business has been driven by abusing the “curiosity gap,” you're now officially in trouble. http://www.nytimes.com/... Josh Lederman / @joshledermanap : I will not click on this. I will not click on this. http://twitter.com/... Ann-Marie Alcntara / @itstheannmarie : CAN FACEBOOK JUST SAY IT'S A MEDIA COMPANY ALREADY http://newsroom.fb.com/... Tom Gara / @tomgara : Pretty harsh that FB sent a robot back in time to kill the Upworthy founder but at least they're putting users first http://newsroom.fb.com/... Seth Fiegerman / @sfiegerman : Facebook: “Websites and Pages who rely on clickbait-style headlines should expect their distribution to decrease.” http://newsroom.fb.com/... @nathanjurgenson : the problem is organizing information and expression based on scores (clicks, likes, etc). people will game it, fb plays whack a mole Alexia Tsotsis / @alexia : Facebook trying to eradicate clickbait is like Malboro trying to cure lung cancer. @mikeisaac : commence existential “do i produce clickbait, dear god i hope not” self-examination by publishers http://www.nytimes.com/... Jake Grovum / @jgrovum : No editorial judgment going on here, Facebook. Nope. http://newsroom.fb.com/... pic.twitter.com/93M4sb7Ns5 See also Mediagazer
Context & Ripple Effects
This is the opening move in Facebook's multi-year campaign to substitute algorithmic judgment for human editors in deciding what news deserves distribution. Within months it escalated from headline quality to truth itself: by December, stories flagged by two or more unpaid third-party checkers like Snopes and AP would get warning labels and lower ranking while staying shareable.
The arc then bent toward legitimacy and structure — regional, non-personalized Trending topics in early 2017, user trust surveys whose design drew public scrutiny in 2018 (whether media could contest their rank), and a boost for local publishers — before the same ranking machinery was tested against publishers directly when Facebook moved non-promoted Page posts to a secondary feed in six countries. This 2016 clickbait rule is where the precedent got set: ranking changes, not policy announcements, became how Facebook governs news.
First-order effects
- Publishers built on curiosity-gap headlines — the 'you won't believe what happened next' format — face an immediate reach cut, since Facebook's classifiers now demote stories that withhold or distort information rather than waiting for user complaints.
- Facebook's ranking team absorbs an editorial function at scale: defining what counts as 'withholding' information is a judgment call previously made by human editors, now encoded in a classifier.
Second-order effects
- Newsrooms retool headline practice toward completeness over curiosity, and some shift effort toward formats the classifier doesn't punish — the observation circulating alongside this story is that killing clickbait is also a reason to show more video instead of articles.
- Each subsequent ranking intervention — fact-check labels, trust surveys, the secondary-feed test — extends the same lever, forcing publishers to optimize for Facebook's stated criteria rather than audience behavior alone.
Third-order effects
- If the pattern holds, news distribution on social platforms consolidates around opaque ranking systems that act as de facto editors — with publishers holding no formal appeal, a gap the 2018 trust-survey debate made explicit.
- The dependency cuts both ways: once news reach is governed by one company's classifiers, that company can later withdraw the surface entirely, which is effectively what Meta did when it reallocated resources away from the News tab in 2022.
The trend: Social platforms are replacing human editorial gatekeeping with ranking algorithms, making each classifier update a de facto publishing policy for the entire news industry.