Illustration: Rebecca Zisser/Axios
Twitter announced Wednesday that it will implement prompts asking users to review harmful or offensive replies before they hit send as a way to encourage healthier conversations on the platform.
Why it matters: Twitter, which had been testing the prompts for almost a year, said the notifications resulted in users sending fewer offensive replies and broadly improved behavior on the platform.
By the numbers: The company noted that throughout testing, 34% of people revised their initial reply or decided not to send a reply at all if prompted.
- On average, users composed 11% fewer offensive replies after being prompted at least once.
The big picture: Social media platforms, including Twitter, have been criticized for years for not doing enough to curb harassment on their platforms even after promising to do so.
Go deeper: Startup backed by Reddit co-founder helps Twitter users block trolls