Something is amiss on Twitter.
Over the last few weeks, Twitter users have reported that a strange glitch in the algorithm has affected the amount of engagement their tweets receive. And it seems to be the opposite way one would assume the algorithm operates.
“Setting your account to private vastly improves your reach by a factor of 5x,” one person tweeted.
“In my week of being private, engagement with my tweets considerably went up about 2x,” another wrote.
Hundreds more tweets like this have appeared on the app over the last month indicating something may be wrong with the algorithm.
Sign up for our free Indy100 weekly newsletter
As Twitter’s “chief complaint hotline operator” and CEO, Elon Musk decided to put it to the test.
On Wednesday afternoon, Musk set his Twitter account to private after declaring “something is wrong” in response to a tweet from LibsofTikTok.
“Made my account private until tomorrow morning to test whether you see my private tweets more than my public ones,” Musk wrote.
Users tracking public data reported that Musk's private tweets were getting more views and engagement than his public ones.
\u201cIt looks like @elonmusk has also proven that the private account Tweets get far more engagement due to something with the algorithm. This matches my test...\n\nLocked (Left) - 3.4 million views in just over 1 hour w/no retweets.\n\nUnlocked (right) - 4.2 million views 9 hours ago.\u201d— Gain of Fauci (@Gain of Fauci) 1675236249
Some theorised why the strange anti-algorithm issue may be true, starting with the fact that by locking your tweets, Twitter must boost them to your current followers since they cannot reach new people.
"My hypothesis is that anyone following you after your account is locked MUST dig your content, by definition. So it's boosted amongst your followers. You can't get new ones, but old ones see more," Ed Latimore wrote.
\u201cnot particularly shocking that private tweets perform better than public in a twitter optimizing for max engagement\n\nyour tweets measured against people who are already opted in, vs your tweets measured against everyone on twitter\n\nmath innit\u201d— @jackbutcher (@@jackbutcher) 1675258194
Other people who have tried out the experiment posted their metrics to prove the theory is true with some asking followers to reply if a tweet had made it to the top of their "for you" or "following" timeline upon going private.
\u201c@elonmusk @libsoftiktok The first was locked, the second unlocked- 20 minute lead in the first.\u201d— Libs of TikTok (@Libs of TikTok) 1675209502
Despite the new theory seemingly working, some people argued against making tweets private for the sake of higher engagement.
People claimed the sudden increase was only short-lived since private tweets cannot be re-tweeted or replied to the same as public tweets.
Have your say in our news democracy. Click the upvote icon at the top of the page to help raise this article through the indy100 rankings.