Recommendation algorithms increasingly shape how users discover content online. Many platforms now rely on social filtering, or prioritizing content from people in a user’s network, to drive engagement and satisfaction. But how does this design choice affect the creation of new content? 

In this webinar, we examine a large natural experiment on a major Q&A platform that shifted from content-based recommendations to social filtering. The results reveal an unexpected tension: while social filtering appears to increase user agreement and satisfaction with content, it may also change how (and how much) users contribute knowledge. 

 We discuss what this paradox means for companies that rely on user-generated content, online communities, and recommendation systems, and how algorithm design can influence both engagement and the long-term vitality of digital platforms.