Debugging and Optimizing User Experience with Deep Visual Analytics
summary
Clarity captures DOM layout, viewport and pointer events, batches them into compressed JSON, and runs ML on those features to detect anomalies and cluster similar sessions. A malware case shows 5% of users affected by injected ads that double error rates; countermeasures restore speed and an experiment yields 10% higher engagement and ~50% faster loads. A food‑blog test adds a jump‑to‑recipe cut abandonment from 30% to 15%. The recorder is open source and a hosted backend is available.
AI-written summary · about
chapters (20)
- 0:00Intro
- 0:27Data Meets Experience
- 3:05Finding purple ads
- 3:45Find a clue in some examples
- 6:56Modular Framework to Instrument the Experience
- 8:22Clarity Analytics
- 10:15Detecting Layout Anomalies
- 11:14Finding Unexpected Content on Bing
- 12:12Debugging Malware
- 13:09Fixing Malware
- 13:38Helps Our Users and Business
- 15:16Food Blag Basics
- 16:36Understanding User Abandonment
- 17:18Finding Similar Abandoned Sessions
- 17:48Finding Coverage Through Similar Sessions
- 18:41Fixing User Abandonment
- 20:08User Delight
- 21:03The confused clicker
- 21:48Confused Clicks on Bing
- 23:11Improving your user experience
related talks
26mDecipher users' behavior using Clarity
16mSquash your hidden bugs using Microsoft Clarity
21mData Analytics for Better Product Decision Making
41m"Analytics and how it should be!"
10mInterview
12mUnlocking the Secrets of Client Data
01h 08mSquashing JavaScript Bugs
46mMass Capture of User Experience with JS