Analyzing digital interactions (webchat, SMS, WhatsApp, and email)

Analyzing digital interactions (webchat, SMS, WhatsApp, and email)

Provisioned feature

The feature described on this page may not be enabled for all accounts. Access depends on your current license level or specific account configuration.

Overview

Conversation Analyzer allows you to process and review digital interactions across multiple channels. By using automated keyword and category detection, you can gain immediate insights into customer sentiment, compliance, and agent performance. You can also re-analyze digital interactions to see the most current data.

The analysis includes:

  • A display of all categories and subcategories linked to the profile.

  • A visual timeline that maps when category matches occurred during the conversation.

  • A full interaction transcript with matched keywords and phrases highlighted.

  • An interaction's metrics, including the times spent from the initial message through queueing and active agent communication.

  • Agent-provided summaries and disposition codes.

Common characteristics:

  • Sentiment analysis is not available for digital interactions.

  • The same categories can be used for analyzing voice and digital interactions.

Prerequisites

To access this feature, you need to:

  • Have Conversation Analyzer enabled for your account.

  • Have a digital channel configured for your account. For more information, see Channels.

  • Configure Conversation Analyzer processing for your agents:

    1. Navigate to User Admin > Agent Settings.

    2. Locate the agent(s) whose digital interactions should be processed.

    3. Set the Conversation Analyzer for chats to Yes. This switch controls all digital channels, including webchat, WhatsApp, email, and SMS.

Analysis components

SS_EmailAnalysis.png

No.

Element

Description

No.

Element

Description

1.

Digital channel

Displays the communication source (email, SMS, WhatsApp, or webchat).

2.

Categorization

Displays all categories linked to the interaction’s profile.

  • Colorful Categories represent categories where matches were found.

  • Gray Categories represent categories with no matches.

Additionally, you can click on a category to hide it. The hidden categories are then highlighted in black.

You can view the subcategory matches by clicking Show subcategories.

The categories are also marked in the Event visualization and Interaction transcript elements.

3.

Conversation timeline

A visual representation of the interaction, indicating which party spoke and when.

Category matches are marked throughout the visualization by colorful dots. You can hover over a category dot to see the specific words or rules that triggered the match.

Playback

A visual representation of the interaction timeline. You can use it to scroll through the Interaction transcript element.

4.

Interaction transcript

Contains the text of the conversation. Words that triggered category matches are highlighted in the corresponding category's color.

5.

IVR Time

The duration from the first message sent until the interaction was assigned to a queue.

Queue Time

The time the interaction spent waiting for an agent to be assigned to it.

Talk Time

The duration from when the agent was assigned until the final message of the session was sent.

Digital channel

Displays the communication source (email, SMS, WhatsApp, or webchat).

6.

Name of the agent(s)

The name(s) of the agent(s) who participated in the interaction.

Disposition code

The disposition code(s) that were entered by the agent(s) at the end of the interaction.

Summary

Agent notes that were entered by the agent(s) at the end of the interaction.

Email limitations

While Conversation Analyzer provides a unified experience, certain behaviors are specific to the email channel:

  • Analyzed Content: Processing is limited to the email Subject Line and Body Text.

  • Excluded Content: Attachments (such as PDFs, Word documents, or images) are not analyzed.

  • Transcript Plaintext: To ensure keyword highlighting is clear and consistent, all original email formatting (bold, italics, etc.) is removed from the transcript view.

Substitution rules

Substitution rules help protect sensitive data, but their application logic varies by channel.

Channel type

Rule application

Channel type

Rule application

Webchat, SMS, WhatsApp

Rules are applied to individual messages in real-time as they occur.

Email Threads

Rules are applied to the entire interaction thread. If a rule is set to "Agent-only," it triggers upon the agent's initial message and propagates through the full thread.

Email limitations

Because email transcripts are processed as HTML documents, there are specific limitations to ensure the document structure remains intact:

  • Avoid HTML Tags: Do not use words that are also HTML tags in your substitution rules. Rules containing these words will be ignored. Examples include: "html", "head", "title", "base", "link", "meta", "style", "script", "noscript", "template", "body".

  • Restricted Templates: Avoid using broad wildcard templates as they can break the transcript structure. The following are ignored for Email: "text *","* text","* text *", "*", "?*", "*?", "*?*".

How to analyze and re-analyze a digital interaction

To analyze a digital interaction:

  1. In the left-hand ribbon, click Interaction Content.

  2. Find the interaction you want to analyze.

  3. Click Open player to view interaction analysis.

To re-analyze a digital interaction:

  1. Click the Re-analyze interaction at the top of the Conversation Viewer.

  2. A banner will confirm that the interaction is being reanalyzed.

  3. Refresh the page to view the updated match results.

Support and documentation feedback

For general assistance, please contact Customer Support.

For help using this documentation, please send an email to docs_feedback@vonage.com. We're happy to hear from you. Your contribution helps everyone at Vonage! Please include the name of the page in your email.