> For the complete documentation index, see [llms.txt](https://library.zoom.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://library.zoom.com/business-services/zoom-contact-center/zoom-customer-experience/cx-insights.md).
# CX Insights
## Overview
**CX Insights** is an AI-native insights experience that unifies structured data (queues, agents, and metrics across WFM, QM, ZVA, and ZCC) with unstructured data (transcripts) to deliver natural-language answers, visual insights, and predictive analysis. Rather than confining users to a single product's dataset, CX Insights spans the full Zoom CX suite, letting leaders and cross-functional stakeholders get answers in as little as a few seconds instead of hours of manual dashboarding and data-stitching. It goes beyond surfacing what happened by extracting context from trends, drivers, and root cause, so teams can understand why.
{% hint style="success" %}
**Example**: A support leader might ask, "Why did escalations increase last week?" and CX Insights could reveal a surge in billing-related conversations following a pricing update as the key driver.
{% endhint %}
CX Insights supports context-aware follow-up questions, allowing users to refine their analysis without starting over. It moves seamlessly from high-level trends to granular segments like agents, queues, flows, dispositions, and time windows. It also automatically generates rich visual artifacts like charts, summaries, and shareable outputs—designed for leadership reviews and storytelling, making the data actionable even for stakeholders without deep contact center expertise.
## CX Insights vs. Traditional Analytics
| Traditional Analytics
| CX Insights
|
| ----------------------------------------------- | ---------------------------------------- |
| Shows what happened | Explains why it happened |
| Fragmented across products | Unified view across CX suite |
| Manual analysis required | AI-driven insights in seconds |
| Dashboard navigation | Conversational interface |
| Reactive reporting | Proactive recommendations |
| Contact Center users only | Leadership and cross-functional personas |
## Licensing and Role Assignment
Before users can access CX Insights, they must be assigned a CX Insights license.
1. Sign in to the Zoom web portal.
2. Navigate to **User Management** → **Users**.
3. To the right of the user’s email or name, click **Edit**.
4. Under Licenses and add-ons, check the **CX Insights Leader** box.
5. Click **Save**.
After the license is assigned, add CX Insights users to the appropriate role:
1. Navigate to **User Management** → **Roles**.
2. Click the **CX Insights** tab.
3. Click **CX Insights Leader (Default)**.
4. Click **Add Members** and add the users as members.
## Language Support
CX Insights natively supports 17 languages:
{% columns %}
{% column %}
* English
* Spanish
* German
* French
* Portuguese
* Simplified Chinese
* Traditional Chinese
* Japanese
* Korean
{% endcolumn %}
{% column %}
* Russian
* Italian
* Vietnamese
* Polish
* Turkish
* Indonesian
* Dutch
* Swedish
{% endcolumn %}
{% endcolumns %}
## Using CX Insights
### Accessing CX Insights
1. Sign in to the Zoom web portal.
2. Click your profile picture or initials in the top-right corner, then select **My Account**.
3. In the side menu, click **CX Insights**.
Don't see CX Insights in your menu? Verify that a CX Insights license is assigned to you, as well as the CX Insights Leader role.
### Configuring Your Settings
1. Sign in to the Zoom web portal.
2. Go to **My Account → CX Insights**.
3. At the bottom of the page, click **Settings**.
From there, you can configure:
* **Time Zone**: Aligns all reports and timestamps with your local area — no more mental math on when that spike actually happened.
* **Collect Queries for Improvement**: A toggle to choose whether your interactions are shared with Zoom to help improve the product.
#### Persistent Memory
CX Insights includes a persistent memory store that retains analytical context across sessions at the user level, offering personalization so users don't need to reset their workspace or re-establish context each time. Memory is captured automatically through repeated actions, user declarations, or saved mappings, and surfaces naturally during conversations. Users can manage all entries, or add new ones, in a dedicated Memory panel under **Settings** → **Personalization**.
### Navigating the Home Page
The CX Insights home page serves as the starting point for exploring what's happening across your CX interactions. Depending on your configuration and permissions, you'll typically find the following areas:
#### CX Pulse
CX Pulse includes key metrics with trend indicators, highlighting whether metrics have increased or decreased within a selected time period. The following metrics are available:
* Engagement Volume
* Service Level
* Average Handling Duration
* Abandoned (%)
* Sentiment
You will also find the **Key Findings & Actions** section, which automatically generates actionable insights ranked by importance. This helps identify any critical issues, why they matter, and even provides additional prompts that will help you dive deeper into findings.
#### Top 10 Operation Spotlight
Rather than requiring managers to build custom reports, the **Top 10 Operations Spotlight** automatically surfaces the top items worth reviewing, turning raw contact center data into a fast, prioritized digest for coaching and operational decisions. Think of this spotlight as the top 10 agents and queues driving performance changes in your contact center.
In **Queues** view, each card shows engagement volume, service level, average handle duration, and sentiment; in **Agents** view, cards instead show engagement volume, average handle duration, adherence score, and evaluation score. Each metric includes a trend indicator, making it easy to spot standout performers, at-risk agents, or unusual shifts at a glance.
#### Top 10 Trending Topics
(Requires a Quality Management license) Topic cards help answer:
* What's showing up **more frequently**?
* What's **newly emerging**?
* What's **declining**?
Selecting a topic opens a deeper view where you can review transcript summaries, narrow the scope further, and select suggested questions without using your own allocated queries.
### Using Natural Language Queries
CX Insights allows leaders to turn customer interactions into actionable data using natural language queries, enabling real-time insights without deep data expertise. Select the **New Chat** tab in the left navigation to start asking questions.
#### What You Can Ask
* **Trending topics** in customer conversations
* **Drivers** behind changes in outcomes
* **Opportunities for improvement** and suggestions
#### Effective Prompt Structure
A reliable prompt pattern is:
**Time range + Scope + Question + Desired breakdown/output**
{% hint style="success" %}
**Examples**:
* "In the last 14 days, what topics are trending up in the Returns queue? Break down by week."
* "In January, what are the main drivers of negative sentiment? Summarize top 5 drivers and suggested actions."
{% endhint %}
#### Tips for Better Results
* **Be explicit about time** — use specific ranges or dates.
* **Name the scope** — queue, team, channel, or region.
* **Ask for a format** — "list top 5," "group by week," "summarize in bullets."
* **Iterate** — start broad, then drill down into specifics.
* **Ask the AI to state its assumptions** — get clarity on scope or metric definitions.
* **Ask for next steps** — request actionable recommendations, not just data.
### CX Insights Projects
CX Insights Projects keep your work organized as usage grows by letting you organize CX Insights chats into persistent, shareable workspaces. You can create new chats within a project or move existing ones in, and you can share a project with teammates for view-only access. Full-text search spans chat titles, questions, report titles, and report content, linking directly to the matching question in a session. A Reports (Artifacts) library at the project level surfaces every generated report in one place, eliminating the need to scroll through chat history.
#### Project-Level Instructions and Memory
CX Insights remembers project-level rules and individual context automatically. Project instructions are shared rules covering default analysis scope, metric definitions, output format, and business preferences that apply consistently across every conversation in a project. Project memory is automatically generated by the system from conversations to retain long-term context, such as a team's shorthand for a queue, a preferred reporting window, or a correction made in a previous session.
### Human-in-the-Loop (HITL) Clarification
CX Insights supports Human-in-the-Loop (HITL) clarification to improve query accuracy and user experience. When the AI encounters ambiguous inputs or lacks sufficient context, it no longer returns a generic error. Instead, it asks a targeted follow-up question directly in the conversation. For example, clarifying between similar entities or confirming an assumed date range. Once clarified, the query resumes with the correct context.
### Data Requirements
For insights that are meaningful rather than anecdotal, Zoom recommends:
* **Minimum 2 months** of historical data
* **10,000+** completed engagements
### Security & Privacy
* **Avoid including sensitive personal data** in prompts unless your organization permits it.
* Treat AI-generated insights as **guidance rather than final judgment** — validate before using them for performance or policy decisions.
* CX Insights uses third-party AI models from providers such as **OpenAI and Anthropic** for certain features.
* Zoom requires all subprocessors to meet obligations **substantially equivalent** to those in Zoom's Data Processing Agreement, and subprocessors undergo **annual security assessments** as part of Zoom's third-party risk management program.
* Security controls are independently audited and included in Zoom's certifications and attestations, available on **Zoom's Trust Center**.
* Subprocessors may retain content for trust and safety purposes within the U.S. or EEA for **up to 30** days, unless a longer period is legally required.
\
\
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://library.zoom.com/business-services/zoom-contact-center/zoom-customer-experience/cx-insights.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.