Improve code quality by testing in production
Testing code directly in production is the surest way to validate your new feature works.
Split makes testing in production possible, with feature flags to control the code, targeting to expose it to a limited set of users, and automated alerting to identify defects.
Why test in production?
Staging is never representative of production. It’s not the same data. And it’s not the same scale. Testing in staging alone doesn’t give development teams the confidence they need to move fast and reduce cycle times.
Testing directly in production gives development teams a clear understanding of how a code change behaves. Pairing this with the ability to limit who sees a change in production means a superior developer experience and a better user experience.
Best practices for testing in production
Get the O’Reilly eBook
Test in production with feature flags
Feature flags allow teams to perform functional QA and performance tests directly on production with a subset of users, providing a secure and performant way of scaling a new feature.
Split gives development teams the power to test in production by targeting new features to QA teams or beta testers and then slowly rolling out to production users.
Should an issue arise, Split alerts you to the feature responsible using data integrations with error tracking and web performance tools. Instantly turn any feature off without having to roll code back.
Test every change where it matters, in production
Feature flags are a framework to safely release code directly to production, where you can really test if it works.