How to think `User first`
summary
Claim: making the site funnier will increase interaction; hypothesis gets tested by a three‑variant A/B experiment implemented with Next.js, feature flags and GrowthBook, shipping analytics to Google Analytics and BigQuery via Firebase. Demo shows brief traffic spikes (reported 200 users per minute), 60 users in one variant with 63 taps, and the awkward‑seal variant trending higher but overall not enough data to declare a winner, so the idea runs but does not conclusively work.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 2:15Why care about users
- 6:10Developers vs users
- 10:32Thai user research
- 14:20User research methods
- 16:53Feature flags
- 19:06A/B tests with metrics
- 21:17Live demo setup
- 29:51Demo results recap
- 30:57Takeaways and call to action
related talks
27mMaking mistakes and building products from 1st to 30M users
52mAB-tests, lies, damned lies, and statistics.
27mA Tale of Two Components: Mastering A:B Testing in React
36mA/B Testing in Production
38mNo Traffic, No Users, No Problem! - Usability Testing for New Apps
41mA/B Testing with Style
35mHow to develop Growth
24mGetting Started with Multi-variant Testing