All flagging and targeting decisions happen in-memory in client, not Split, servers in a few microseconds. Thus, there’s no noticeable latency or lag when serving flags to your end users, ensuring a seamless user experience.

Split is built from the ground up with privacy in mind. When targeting on private attributes, such as email or demographics, our SDKs do the heavy lifting. Sensitive user data stays within your app or server and no private data (PII) is sent to Split. No user identifiable data is retained by default – as sharing sensitive user data to our cloud service is not required to target features.

Split servers are not in the critical path of your application. All flags and rules are cached locally in our SDK, so your application continues to perform as expected even in the case of a CDN outage. Split data and servers are replicated in multiple regions to guarantee uptime and availability of our flags. Learn more about reliability

Split is designed to make every change easy to reason about. We record every flag evaluation for every user — meaning you can see which flag was served to which user at any time, giving you unparalleled debugging capabilities. We also track usage in terms of impressions, events, and monthly unique IDs across all environments and across all workspaces.

Secure access to Split with SAML-based single sign-on and two-factor authentication, and verify the authentication status of any user at any time. Easily manage large engineering teams with role-based access controls, managed via groups, that allows administrators to set permissions by-environment and by-feature. Split is SOC 2 certified to support enterprise environments. Learn more about security