The Latest in Metrics & Measurement
summary
Defines and compares concrete metrics and collection tools: First Contentful Paint as first non‑white paint, First Meaningful Paint waiting for webfonts (except tiny icon fonts), and Speed Index computed as the integral of visual completeness via color‑palette comparison and now reported in milliseconds. Explains interactivity thresholds — 50 milliseconds marks a long task, 100 ms target for input latency, 95th percentile for FID — and describes field tooling: PerformanceObserver, Long Tasks (with privacy/attribution limits), Lighthouse, WebPageTest, Perfume.js and user‑timing instrumentation.
AI-written summary · about
chapters (29)
- 0:00Intro
- 0:55Load
- 2:00Visual
- 3:00First meaningful paint
- 7:47Speed index
- 13:11Other considerations
- 13:51Life advice
- 19:36Main thread tasks
- 21:21Rail model
- 22:19Input latency
- 25:04Smaller tasks
- 25:31Time to Interactive
- 27:31First CPU Idle
- 28:44First Input Delay
- 30:11First Input Late
- 31:11Expected Queue Time
- 33:51Lab Data
- 35:46Field Data
- 35:59Long Tasks
- 37:32Paint Time
- 38:24Perfumes
- 41:53Closing Keynote
- 42:38TTI
- 44:06First Content Full Paint
- 46:35CPU impact
- 47:51Page ranking metrics
- 49:15User timing spec
- 51:14Hero rendering times
- 52:33Setting up an observer
related talks
30mMeasuring Web Performance - Metrics & Tools
36mMoneyball for Performance Metrics
01h 01mUnveiling the Power of Web Performance Metrics for Testers
48mFrom Here to There
21mState of the Union for Speed Tooling
20mReal metrics for front-end performance
35mWeb Application Performance Testing from Users Perspective
37mThe First Meaningful Paint