Project Context
Meta's Facebook Android team is preparing a performance improvement launch for a Feed surface with complex list cells, autoplay video, and heavy image loading. Recent profiling shows repeated main-thread stalls during fast scroll, causing dropped frames and visible jank on mid-tier Android devices. You are the mobile engineer leading execution across 8 engineers, 1 product manager, 1 designer, and shared support from the Infra Performance team.
The project has urgency because the VP of Mobile has asked for a launch plan before the next app release train in 10 weeks. Feed is a top engagement surface, and internal metrics show that sessions on affected devices have a 4.2% lower scroll depth and higher abandonment.
Key Stakeholders
The Facebook Feed PM wants a visible improvement in user experience this quarter. The Engineering Manager wants a low-risk rollout that does not destabilize image delivery or ranking-related surfaces. The Infra Performance team prefers reusable platform fixes, while the Feed feature team wants targeted optimizations that can ship faster. Release Engineering requires a clear rollback plan before enabling the change to more than 25% of Android users.
Constraints
- Timeline: 10 weeks to land in the next Android release train
- Budget: no new headcount; only 2 engineer-weeks available from Infra Performance
- Devices in scope: Android devices with 4 GB RAM or less, representing 38% of Feed sessions
- Current baseline: 11.8% frozen frames and P95 frame render time of 42 ms on target devices
- Dependency: image pipeline changes require review from the Fresco maintainers by Week 5
Complications
- A parallel Feed ranking launch is already scheduled for Week 8 and cannot slip.
- One proposed optimization reduces image quality for slower networks, which the PM worries could hurt content engagement.
- iOS has a similar issue, and leadership may ask whether your approach should be cross-platform.
Deliverables
- Create a 10-week execution plan to reduce main-thread blocking in Feed list rendering and image loading.
- Define the scope trade-offs between reusable infrastructure work and Feed-specific fixes.
- Propose a launch and rollback strategy, including staged rollout percentages and monitoring.
- Identify the top execution risks, dependencies, and escalation points.
- Define measurable success criteria for performance and user impact.
