Best Practices for Feature Flags, Phased Rollouts, and A/B Tests
How to get started with feature flagging and experimentation safely, quickly, and at scale
It’s no longer enough to simply move fast when it comes to software development teams. Product and engineering teams have to reduce uncertainty, make data-driven decisions, and deliver the right features to end users more quickly to drive retention and revenue.
Leading teams have adopted the twin processes of progressive delivery and experimentation to test and learn. However, incorporating any new processes or platforms can bring up many questions for engineering teams.
In this how-to guide, we try to answer those questions, along with:
- How to get started with feature flagging and a/b testing without creating long-term technical debt
- How to scale to thousands of flags with good governance and reliable QA processes
- How to adopt these new practices organization-wide without slowing down feature development and delivery
