Hi HN,
At several companies I worked at, feature flags followed the same pattern:
Someone suggests LaunchDarkly. Everyone agrees it’s solid. Then someone sees the pricing. The team decides to “just build something simple.”
A few months later:
- Flags scattered across the codebase - No consistent rollout logic - No structured targeting - Weak environment separation - No proper kill switch - Old flags never cleaned up
Rolling your own starts simple but slowly becomes infrastructure you didn’t intend to maintain.
After seeing this happen more than once, I built FeatureFlare:
The goal isn’t to compete with enterprise platforms. It’s to serve small SaaS teams (1–10 engineers) that want proper feature flagging without sales calls or enterprise overhead.
Core functionality:
- Per-environment flags - Percentage rollouts - Rule-based targeting - Immediate kill switches - Simple API and SDK usage
The focus is on keeping flag logic centralized and structured so it doesn’t turn into conditionals scattered across the codebase.
I’d really appreciate feedback on:
- Is this actually a meaningful gap? - Do most small teams just roll their own? - What would make you not build this internally?
Happy to answer any technical questions about architecture, scaling model, or tradeoffs.