Dialog Analytics
Overview and techniques for analyzing dialog interactions
Last updated
Overview and techniques for analyzing dialog interactions
Last updated
Collecting usage metrics of your project is essential to improving the dialog and can provide useful insights about how users interact with it. You can do so by connecting your release to an analytics provider of your choice.
When creating a release, you will be prompted to specify an endpoint for receiving Analytics data (referred to as the Analytics URL). Parloa will collect details from every conversation with your project and send them in batches to the designated analytics hook endpoint.
From Parloa's perspective, providing the URL of the endpoint is the only requirement. Should the endpoint be protected, credentials may be supplied through the Authentication parameter or by specifying a custom header.
You can test that the connection between Parloa and the analytics hook is working by clicking the "Send Test Request" button. This will send an HTTP POST request with an empty JSON body and the following header:
The analytics hook is expected to reply with an HTTP Status 200 in order to acknowledge the test request.
