Driving App Stability & Performance Zero to Hero
summary
Defined a single Application Quality Score (AQS) that combines seven sub-metrics (crash‑free users, ANRs, frozen frames, slow rendering, app start-to-interactive, core screen TTI, app size) with stability weighted ~60% and targets like 99.9% ideal crash‑free users. Engineering actions: reduced third‑party SDK impact, fixed Jetpack Compose textfield crashes, cut shared‑preference fsyncs, deferred heavy init with lazy/provider dagger, and profiled startup. Tooling: Firebase Performance, Sentry, Regresso for traces and regression alerts; outcome = AQS 0 → 70 in one year with 200+ remediation tasks completed.
AI-written summary · about
chapters (22)
- 0:00Mobile app quality intro
- 1:45Defining technical quality
- 4:15Why quality matters
- 4:37Introducing the Application Quality Score
- 5:19How AQS is calculated
- 8:25Metrics weights impact
- 10:31Setting AQS targets
- 12:41Initial AQS score wake-up
- 13:47Three-phase AQS strategy
- 15:55Crash fixes deep dive
- 18:02Jetpack Compose text field
- 21:54Understanding ANRs
- 24:20ANR ripple effect
- 27:08Security vs performance
- 29:17Firebase ANRs & SharedPrefs
- 34:11App start time improvement
- 37:02Async init for startup
- 38:51DI defer for performance
- 39:54Fix frozen frames
- 41:18AQS results after 1 year
- 42:00Maintaining stability
- 43:25Key takeaways & future
related talks
30mBoosting Performance Through App Quality Improvements
39mAndroid App Performance in a Nutshell
41mModern App Performance
19mBeyond the Basics: Performance Monitoring and User Experience for Mobile App Growth
40mAndroid vitals: Debug app performance and reap rewards
32mChasing performance issues methodically
40mAttributed Everything: Ensuring Quality for Large Apps
30mLearn from developers: Lyft and Google Maps share their experiences optimizing app performance