A/B testing and canary releases in 2022 - why, what, and how?
The engineer's perspective on experimenting and finding a cure for scurvy 🚀
Having proper options for testing, collecting qualitative and quantitive data, and understanding it is crucial to success in product development. A/B tests and canary releases are separate topics that serve different purposes, but as they have some overlapping parts in terms of organizing your development process and code architecture, it might be worth taking a look at both of them. In this session I'll:
- elaborate on the need for those solutions, go through a few successful (and disastrous) case studies from my experience,
- talk about tools that can help you efficiently manage, optimize and measure the process,
- give hints on how to tackle the problem of insufficient data from the statistical standpoint and other pitfalls you should know about,
- say a few words about ideas for tests and where to look for knowledge in this area,
- tell you a story of finding a cure for scurvy - and failing on the way.
A business-driven engineer with a deep passion for frontend (especially web performance) and product development. Apart from regular work, Rafal actively contributes to the community by organizing local meetups (Angular Dragons) and mentorship (girls.js/ngGirls workshops). In his free time, he enjoys diving into areas of psychology, history, or any other field that got to his heart.