Web Application Performance

The Gainsight PX Tag 

Tracking in-product usage requires installing a Javascript code snippet AKA Gainsight PX Tag, that can collect user interaction and send it to Gainsight PX tracking and personalization service.

Designed For Browser Resource Optimization  

• Asynchronous Tracking - the PX tracking code uses non-blocking, async calls to avoid unnecessary latency from any application service used by the end-user.
• Concurrent Connections - browsers have limited resources that are designed to serve the application's front-end by fetching the relevant data, code, and styling. Gainsight PX uses a single concurrent connection for tracking purposes to avoid resource hogging.
• In-Product Engagements -  most in-product engagement solutions are using the browser to cache ahead in-product engagements when serving 'onboarding widgets' or 'knowledge centers' that may contain on-demand product tours. This leads to application page latency and resource-hogging since the browser is required to download and store all the relevant engagements even if the user never triggers any of them. Gainsight PX uses server-side caching to avoid that browser resource-hogging scenario.
• CDN Service - Gainsight PX is using Google's Content-Delivery-Network to optimize the SDK serving. Once the SDK is loaded it will wait for the page to complete loading before tracking or serving in-product guides.

Server-Side Architecture 

Gainsight PX's real-time personalization engine is using a distributed micro-services architecture with auto-scaling (K8s). The server-side latency for any in-product engagement campaign evaluation is executed in parallel with an average latency of 1-5 milliseconds.