Kris Jenner, Kendall Jenner, Kylie Jenner and Kim Kardashian on May 8, 2018 in New York City.
Dimitrios Kambouris | Getty Images Entertainment | Getty Images
Meta is ending a test version of its Instagram app that earned the ire of many people, including celebrity sisters Kylie Jenner and Kim Kardashian.
Instagram chief Adam Mosseri told the Platformer newsletter on Thursday that it would roll back its recently released version of Instagram that predominantly featured algorithmically recommended videos to users on their home screens.
Numerous Instragram users, including the Kardashian sisters, complained about the test version of Instagram, voicing frustration that the app was overtly copying rival TikTok while distancing itself from its photo-sharing roots.
Mosseri then released a short video on Tuesday responding to critics in which he said that photos will continue playing an important role for Instagram, but changing user habits favor video. In other words, Instagram will continue pushing into video because people are increasingly sharing and liking videos on the app, he explained.
But apparently the backlash against the test version of a video-focused Instagram led to Mosseri and his team temporarily reversing course.
“I’m glad we took a risk — if we’re not failing every once in a while, we’re not thinking big enough or bold enough,” Mosseri told Platformer. “But we definitely need to take a big step back and regroup.”
Still, Mosseri made clear that Instagram will continue to increasingly cater to video, and that while it will temporarily reduce the number of algorithmically recommended videos people see, that number will rise again once the company feels its improved the technology.
When Meta reported second-quarter earnings that missed Wall Street expectations, the company’s CEO Mark Zuckerberg told analysts during a call that he “about 15% of content in a person’s Facebook feed and a little more than that in their Instagram feed is recommended by our AI or accounts that you don’t follow.” Meta expects those numbers “to more than double by the end of next year,” he added, underscoring the company’s reliance on using machine learning to figure out what material to show users.