How we think about Threads’ iOS performance

This post details how the Threads iOS team approaches performance monitoring and optimization. Key contributions include: defining and monitoring metrics like %FIRE (Frustrating Image-Render Experience) and TTNC (Time-to-Network Content) to track image loading and feed delivery speed, and cPSR (Creation-Publish Success Rate) for post reliability. It introduces SLATE, a custom 'Systemic LATEncy' logger, for consistent measurement of navigation latency across various user interactions and app states. The post also highlights the importance of keeping the app's binary size small, citing a 75% reduction for Threads compared to Instagram, and faster build times. A case study on adopting GraphQL for network requests is presented, emphasizing the need for performance validation using SLATE before full adoption. The team also conducted a 'boundary test' by injecting latency to understand user tolerance and impact on engagement.