Knowledge Insights for Solve Email

Knowledge Insights for Solve Email provide valuable data on how your knowledge base articles perform within Solve Email. By analyzing these insights, you can optimize your articles to deliver more accurate and higher-quality email support.

Where to find it:

To see Knowledge Insights, go to Solve > Solve Insights > Knowledge tab > Email.

Key Benefits

• Identify High-Performing Articles: Discover which articles contribute to successful email resolutions.
• Pinpoint Content Gaps: Recognize areas where your knowledge base lacks necessary information.
• Enhance Article Quality: Optimize articles to improve response accuracy and resolution rates.

Use Case

Consider a support team that often receives emails regarding password resets. They have an article in their knowledge base about password resets, which automatically surfaced within email responses. When they reviewed their Knowledge Insights, the team noticed a significant amount of negative feedback associated with this article. The team recognized the issue, so they revised the article by providing clearer and more concise steps. As a result, end-users were able to successfully reset their passwords, which led to positive feedback and fewer tickets reaching human agents.

Emails sent with articles surfaced

In the Email tab, you'll find the "Emails sent with articles surfaced" card. This card provides insights into the number of emails sent where knowledge base articles were presented to the sender. It also tracks the number of emails that resulted in deflection versus non-deflection.

Definitions:

• Deflection: An email is considered deflected when Solve Email successfully resolves the sender's issue without requiring intervention from a human support agent. A high deflection rate means that the articles have been really helpful in deflecting emails.
• Non-Deflection: An email conversation results in non-deflection when Solve Email, despite surfacing relevant knowledge base articles, cannot resolve the issue and requires human agent intervention. A high non-deflection rate means that you need to improve your knowledge base articles so that your end-users will be able to resolve their issues independently.

User feedback when article was surfaced

In addition to the "Emails sent with articles surfaced" card, you'll find the "User feedback when article was surfaced" card. This card displays the number of times end-users indicate whether the email with surfaced articles was helpful or unhelpful by clicking the thumbs-up or thumbs-down buttons at the end of the Solve Email. 

• Positive: The end-user clicked the thumbs-up button, indicating that the email with the surfaced article was helpful in resolving their issue.
• Negative: The end-user clicked the thumbs-down button, indicating that the email with the surfaced article was not helpful in resolving their issue.
• Not Answered: The end-user did not click either the thumbs-up or thumbs-down button.

Knowledge 

The Knowledge provides key metrics about your knowledge base articles. It contains the following:

• Title: The name of the article.
• Article: A preview of the article's content. Click this to view more information about the article.
• Conversations: The number of email conversations in which the article was surfaced.
• Deflections: The number of email conversations deflected where this article was surfaced.
• Surfaced: The total number of times the article was surfaced in email conversations. 
• User Feedback: User ratings (positive/negative) of the email containing the surfaced article.
• Top Surfaced Workflows: The workflows most frequently associated with this article.
• Related Articles: Articles that are related or similar to this article.

Conversations tab

Clicking an article in the Knowledge table redirects you to the Conversations tab, which provides detailed information about specific email conversations where the selected article was surfaced. The Conversations tab features a table with the following columns:

1. Timestamp: The date and time of the email conversation.
2. Ticket ID: The ticket ID associated with the email.
3. Incoming Email: The email containing the surfaced article. Click to view the full email conversation.
4. Deflected: Indicates whether the email was deflected.
5. Eligible: Indicates whether Email Automation rules allowed Solve Email to respond.
6. Email Sent: Indicates whether an email containing the surfaced article was sent.
7. No response reason: Explains why the email was not answered. Possible reasons include:
   • No Value: No intent was confidently predicted, or the workflow ended without a response.
   • Email Automation: An automation rule prevented a response.
   • Non-Email Channel: The ticket originated from an unsupported channel (currently supports Email, Form, and API).
   • Responded to by Agent: An agent responded before Solve Email could.
   • Ticket Closed: The ticket was closed before Solve Email could respond.
   • From Widget Handoff: The ticket was created by the Solve Widget.
   • No Intent Predicted: An active intent could not be predicted, and Knowledge Retrieval was not activated.
   • Workflow Execution Error: An error occurred during workflow execution, often due to an Action API call or missing Context Variable.
8. Detected Workflow: The workflow detected in the email conversation.
9. Surfaced Articles: The articles surfaced in the email conversation.
10. User Feedback: User ratings (positive/negative) of the email containing the surfaced article.

Related Articles tab

The Related Articles tab displays information about similar articles related to the selected article. It presents a table with the following information about each related article:

• Title: The name of the article.
• Article: A preview of the article's content. Click this to view more information about the article.
• Conversations: The number of email conversations in which the article was surfaced.
• Deflections: The number of email conversations deflected where this article was surfaced.
• Surfaced: The total number of times the article was surfaced in email conversations. 
• User Feedback: User ratings (positive/negative) of the email containing the surfaced article.
• Top Surfaced Workflows: The workflows most frequently associated with this article.
• Related Articles: Articles that are related or similar to this article.

Knowledge Gap 

Our AI identifies content gaps within your knowledge base based on analysis of support ticket interactions. It then generates relevant and useful articles to address these gaps. To add a generated article to your knowledge base:

1. Click a generated article.
   

   

2. Review the article's content.
   
3. To upload directly to your knowledge base, click Upload. To make changes first, click Copy and paste the content into your preferred online document editor.
4. When you click Upload, select the connected knowledge base where you wish to upload the article.
   
5. Choose the specific section within that knowledge base.
6. Once you are done, click Upload as a draft.