Frontend and analytics
summary
Shows how to collect real browser metrics by using the ResourceTiming API, sending resource lists to Yandex.Metrica via visit parameters, and exporting raw data through the Metrica API. Describes finding unexpected third‑party hosts and injected scripts from extensions/antivirus, recommends using Content-Security-Policy with report-uri (noting CSP can’t stop privileged extensions), and explains capturing window.onerror, measuring image loads with duration percentiles (median 543 ms, 80th ≈3 s), and running A/B tests with proper segmentation.
AI-written summary · about
chapters (9)
- 0:00Welcome
- 1:54Page resource analysis
- 7:51Third-party scripts and security
- 10:59Client-side error collection
- 13:24Mobile connection analysis
- 18:05Technical and product metrics
- 20:54Experiments
- 25:26Data analysis and machine learning
- 31:56Q&A
related talks
20mReal metrics for front-end performance
59mObservable Web Applications
11mAre They Really Using It? Monitoring Digital Experience To Determine Feature Effect
33mDemystifying Web Performance Tooling
20mBaby Steps for Metrics
06mMeasurement is Not Enough: Rise of Performance Analytics
48mHow I learned to stop worrying and love UX metrics
31mIt’s My (Third) Party and I'll Cry If I Want To