Skip to main content
Real time

Update campaigns and feature flag configurations instantaneously with Real-Time Streaming Architecture

April 28, 2022
Reading time: 
5 mins

Experimenters want to provide customers with the most relevant and personalized experiences possible.

However, the standard method to relay data between the client and server—called “polling”— can create lags between when users encounter tests and when that data feeds into their experimentation platform.

Polling, by its nature, has this delay and it’s up to the experimenter to decide how often to ping the server for requests at regular, fixed intervals.


Depending on the use case, polling can either be an attractive or a costly trade-off:

  • For feature flags, infrequent polling poses the risk of different parts of the customer application having different flag states. This is not an ideal customer experience. However, infrequent polling is less likely to be problematic for experiments because they are not updated as often during runtime.
  • Polling at a high frequency, on the other hand, will cost in network bandwidth, battery and data usage, especially on mobile devices.


Kameleoon is now the first A/B testing platform to offer real-time streaming, a type of architecture where the server pushes information to connected clients whenever content has changed. No delay.

Real-time streaming architecture is a more efficient and timely methodology for experiments and features flag configurations where instant, real-time updates are crucial.  Benefits include:

  • Fast propagation or no latency propagation of new updates
  • Optimized for mobile applications: each client receives updates only when necessary which means less impact on battery life and data usage.


We are making Real-Time Streaming Architecture available for free for beta customers. Please let us know if you’d like to participate in our Real-Time Streaming Architecture Beta program.

Ready to get started? See this step-by-step guide and technical documentation.

Request a demo to see this feature and the Kameleoon platform in action.

Topics covered by this article