Edit

Collecting user feedback in native calling scenarios

  • Article

Introduction

As a third-party library, we must acknowledge that diagnostic data, crucial for troubleshooting, resides on end user mobile devices. Our support model hinges on a symbiotic relationship: to assist your customers effectively, we must ensure you're equipped to do so. This documentation outlines the methodologies to collect diagnostic data when a user reports an issue.

Key development concerns

  • Timely Collection: Collection logs immediately after an issue occurs.
  • Minimize User Effort: Ensure the user can easily report errors within your application.
  • Privacy and Security: Store your user's information securely, engage for consent pro-actively before transmitting data.

Important information to include with User Feedback

Call IDs

Every Call made with the Calling SDK has a Call ID. Call IDs can be used internally at Microsoft to diagnose issues with your call. Collecting Call IDs is the frontline of support, and can lead to faster investigations.

In both the Calling Native and UI SDK, APIs exist to retrieve Call IDs.

Log files

The Native Calling SDK and its dependencies output encrypted .blog files into a temporary directory. These files can't be read outside of Microsoft. They're encrypted for privacy and compliance reasons. These files are the source of truth as to what is happening on that particular device with the Native SDK and its dependencies. These files are an important resource to developers and troubleshooters within Microsoft.

Scenarios requiring .blog files should be more rare. These files form an important second line of defense for troubleshooting. We encourage developers to proactively collect them via your support flows.

Enabling feedback from the customer

Integrating user feedback tools

Once you know what data to collect, you need to satisfy the following user story.

As a user, I would like to report an issue

Each application is free to implement user facing support in whatever way best fits the use case. Alternatively for UI SDK Users the built-in mechanism is available to help partially satisfy the story.

  • Report an Issue Form: A button and a form, click to submit. The UI Library offers a ready-to-use implementation of this form.
  • End-of-Call Feedback: Solicit feedback at the end of a call. The feedback form gives the user the opportunity to share issues they had with the call.

It's crucial to design these feedback mechanisms with clear prompts for user consent, ensuring users are fully informed about the data being shared and its purpose. This transparency builds trust and encourages more users to report issues.

Sending support information and feedback to a server

Transmitting the support information

Once feedback is collected locally, it needs to be submitted to a server. Submission would typically target a CRM (Customer Relationship Management) or other tooling that can handle tasks like triaging, prioritizing, and assigning work to support specialists.

While this document isn't meant to cover the entire premise of client/server communications and all the possible CRMs or Support tools out there, note the following points:

  • Use secure transmission protocols
  • Include the Logs and the Call IDs when creating support requests
  • Include any User submitted information (Message, Time of error, Device specifications)
  • Provide the User follow up information for their issue (Options: Notify in App, Email, Text Message)

The application developer is free to decide how to transmit this data as it leaves the end users device and enters a server on the cloud. For a simplified example you can reference the collecting user feedback tutorial, which offers insight into both client and server implementations of this process.

Implementing in Calling SDK and UI Library applications

For developers utilizing the Calling SDK or the ACS UI Library, consider the following tools and APIs:

Conclusion

This guide emphasizes the necessity of efficiently collecting and transmitting diagnostic data for effective user support in native calling applications. Employing these strategies enables the rapid diagnosis and resolution of user issues, improving the app's performance and user satisfaction.