Meta Reels Recommendations Improve With User Feedback Surveys
Meta says Reels recommendations improve after it changed how feedback shapes content suggestions. The company now uses direct user surveys inside the Reels feed. As a result, recommendations better reflect real interests. Many users have seen these prompts already. They appear between videos and ask how a Reel made someone feel. Meta rolled out these surveys at scale to gather clearer signals. Previously, recommendations relied on likes and watch time. However, those signals do not always show true interest. Therefore, Meta added survey responses to balance the system. According to Meta, this shift works. Alignment with real user preferences jumped from 48.3% to over 70%. As a result, Reels now feel more personal for many viewers.
How Feedback Shapes Reels
Meta adjusts survey data to reduce bias. It also corrects gaps from people who skip responses. In addition, machine learning blends this feedback with existing signals. This approach helps serve users with limited engagement history. It also supports more diverse recommendations. However, Meta admits more improvements remain.
Other platforms use similar methods. For example, Pinterest has long relied on surveys to refine recommendations. Still, Meta’s progress stands out.
Why TikTok Still Leads
Despite gains, TikTok remains the benchmark. Its For You feed keeps users scrolling longer. That edge likely comes from stronger visual recognition.
TikTok can detect detailed elements inside videos. These include objects, settings, and physical traits. As a result, it matches content more precisely.
Meta lacks that depth today. Instead, it depends on engagement data and surveys. Therefore, Reels recommendations improve, but they still trail TikTok. Even so, creators may benefit. Better matching means content reaches more interested viewers. Over time, that could boost meaningful engagement.
