Platformer, a newsletter on the intersection of Silicon Valley and democracy, recently discussed the ongoing challenges faced by Instagram and its transition towards algorithmic recommendations. In July last year, the platform faced a crisis as its algorithm began to fill users’ feeds with recommended content instead of posts from friends and influencers they had chosen to follow. This change prompted widespread backlash, with many calling for Instagram to “Make Instagram Instagram again.” However, over time, users have become accustomed to the new recommendations and the uproar has mostly subsided.
The transition away from Facebook’s old friends and family-dominated feeds to Meta’s algorithmic wonderland, while relatively smooth, has presented new challenges for the company. To recommend posts effectively, Instagram needs to have a thorough understanding of its interconnected systems and be able to explain to users why they are seeing particular content. Without this understanding, harmful content could be promoted, and users may have questions about the rationale behind the recommendations they receive.
In response to these challenges, Meta, Instagram’s parent company, has published 22 “system cards” that explain how content is sourced, the signals used to make predictions, and how posts are ranked in users’ feeds. The aim is to give users a sense of control over their experiences on the platform and address concerns about Meta’s role in shaping their feeds. Meta is also introducing a “Why am I seeing this?” feature to Reels on Facebook and Instagram’s explore page, further empowering individual users.
While the system cards and transparency efforts may not surprise users who are familiar with social media, their publication could fuel new critiques of Meta. Some may argue that such transparency simply reveals how Meta designs its apps to be addictive, as the system cards outline factors like session length, predicted viewing time, and engagement levels. However, Nick Clegg, Meta’s president of global affairs, defends ranking content based on engagement, comparing it to newspapers or authors choosing stories that readers are likely to enjoy. He emphasizes that Meta also takes into account slower time signals, overall user satisfaction, and feedback surveys.
Related Post
Meta is also testing a new feature that allows users to mark their interest in a recommended Reel explicitly. This feedback may lead to better-tailored feeds and improved user experiences. Additionally, Meta is expanding its collaboration with academic researchers to gain insights into its systems and improve transparency further. The company has announced that it will make a library of public posts, pages, groups, and events on Facebook available to qualified research institutions through an application process.
Overall, while transparency efforts may not solve all the challenges associated with algorithmic recommendations, they are an important step toward understanding and improving the present. By providing users with explanations and involving academic researchers, Meta aims to build a better future for its platforms. Despite potential criticisms and concerns, transparency remains a valuable tool that can foster trust between users and social media companies.
