# Using Data Explorer

> Analyze your Kustomer data with natural language. Data Explorer explains trends, surfaces drivers and anomalies, and suggests next actions from Reporting.

Source: https://support.kustomer.com/en_us/using-data-explorer-HJB0g5wyl

Last updated: 2026-10-23T15:59:52.659Z

Kustomer's Data Explorer is your natural language analysis workspace for Kustomer data. It runs validated queries across your data, interprets it, and returns explanations, patterns, and actionable recommendations you can instantly operationalize.

Click here to learn more about the [Query capabilities of Data Explorer](https://help.kustomer.com/en_us/understanding-the-query-capabilities-of-data-explorer-B1wuEAFcMe).

### Key Features

Traditional reporting explains "what happened." Kustomer's Data Explorer goes further to answer "why" and provide actionable insights.

*   Highlights drivers, outliers, and anomalies that affect staffing, service quality, and efficiency.
*   Explains operational significance in plain language, and prompts managers to act on specific data points.
*   Produces structured results for reuse in dashboards or reviews. Simply copy and paste.

### In this article

*   [Using Data Explorer](#using-data-explorer)
*   [Using Data Explorer's plan mode](#plan-mode)
*   [Accessing Data Explorer](#accessing-data-explorer)
*   [Getting started](#getting-started)
*   [Analyzing data in Data Explorer](#analyzing-data-in-data-explorer)
*   [Prompting best practices](#prompting-best-practices)
*   [Troubleshooting and FAQs](#troubleshooting-and-faqs)
*   [Upcoming enhancements](#upcoming-enhancements)

### Using Data Explorer

Ask in natural language and get analysis across common operations and CX scenarios, such as:

*   Rep scorecards, coaching recommendations, and best practices
*   Capacity and backlog analysis, including queue pressure and aging
*   Benchmarking across teams, channels, or time periods
*   Forecasting and predictive modeling for volume or handle time
*   Voice of customer insights using sentiment and satisfaction data

Tip: Use the **Prompts** menu to browse curated templates organized by analysis category. Templates are a fast way to get high quality results without starting from a blank input.

### Using Data Explorer's plan mode

Instead of immediately attempting to answer a complex reporting request, Data Explorer's plan mode breaks the request into smaller steps. You can review these individual steps, make changes, or request the report dive deeper with greater insight into what the Architect intends to do before it begins work. See Understanding plan mode for use cases and more detail.

With plan mode, outputs are:

*   Reviewable and approvable
*   Assembled with clear intent
*   Complex and detailed

Access plan mode from the **Reporting** page while working with the **Data Explore** skill.

To access **Data Explorer** and use plan mode:

1.  Open **Reporting and use Data Explorer**.
2.  Ask a reporting question that requires planning, decomposition, or multiple reporting steps.
3.  Review the Architect's proposed plan.
4.  Approve the plan to execute it.

Plan mode lists the steps the skill will take to complete your objective, including tool calls it will use to pull data for review. It will explain where the data comes from and why it intends to use that source. Once you've run plan mode, you can execute the plan directly or review it and ask for changes. Enter a follow up message explaining what you need changed or added to the plan, and approve when you're ready.

#### Using chart drilldowns in Data Explorer

Use drilldowns to inspect the records behind a generated chart. This helps confirm that the chart reflects the expected filters and categories.

To use a chart drilldown:

1.  In _Data Explorer_, run a prompt that returns a supported chart.
2.  In the generated chart, click a bar, slice, or other supported chart segment.
3.  Review the filtered result that opens.
4.  Confirm that the returned records match the chart segment you selected.

Drilldowns are most useful when a user needs to answer questions such as:

*   Which records make up this count?
*   Which items are included in this category?
*   Does this segment match the expected date range or filter context?

#### Saving supported chart output

Some generated chart output can be saved and reused. This helps users keep useful analysis instead of recreating it from scratch.

To save supported chart output:

1.  In _Data Explorer_, run a prompt that returns a supported chart.
2.  Review the generated chart and confirm that the result is correct.
3.  Click **Save Analysis as Report** to save the generated result.
4.  Open the saved result and confirm that it renders as expected.

Saved output is intended to preserve useful analysis in a reusable form. It is best to verify the generated chart before saving it.

### Accessing Data Explorer

AI-enabled features must be toggled on for your organization. An admin can do so, or verify if this setting is enabled.

To toggle AI-enabled features for your organization:

1.  In the left navigation panel, click **Settings**. The **Settings** page opens.
2.  In the **Settings** panel, navigate to **Administration**, and then click **Organization**. The **Organization** page opens.
3.  At the bottom of the **Organization** screen, toggle **AI-Enabled** features on or check to verify the setting is enabled.

Note: A prompt featuring Data Explorer greets users upon first time use. It is possible for admin users to toggle AI-enabled features within this prompt.

With AI-enabled features toggled on, access Data Explorer:

1.  In the left navigation, select **Reporting**. The **Reporting** screen opens.
2.  At the top of the **Reporting** screen, select **Data Explorer.**
3.  **Prompt**Data Explorer:
    1.  Use the **Prompts** button to open analysis templates by category.
    2.  In the center of the screen, select from any of the pre-defined prompts.
    3.  Type a question into the text field.

Note: Users can also speak to the bot. Click **Start Voice Input** and speak into your microphone. As you speak, your voice is translated to text in the text input field.

### Getting started

![](https://cdn.kustomerhostedcontent.com/media/693203f25d9f06218fbf4b3d/c733dac2fd18e0d3971d4b69db06b710.png)

1.  Open Data Explorer. You will see suggested actions that run immediately on your data.
2.  Pick a starting template from **Prompts** or type your own request in natural language.
3.  Review the result. Each answer includes the analysis being formed, filters applied, and a detailed explanation of the operational significance.
4.  Iterate. Ask follow-ups like "break that down by channel" or "trend this by week for the last 12 months."
5.  Export your results to PDF or copy the URL directly to share the active thread with a colleague

### Analyzing data in Data Explorer

Data Explorer understands the core entities in your Kustomer workspace and their relationships. Data Explorer can use data from these Kustomer data sets:

**Category**

**Metrics/Details**

**Conversations**

Such as:

*   Channel
*   CSAT
*   Response times
*   Sentiment
*   Status
*   Tags
*   Volume |  
    | **Customers** | Such as:
*   Activity history
*   Conversation counts
*   Engagement
*   Locations
*   Profiles
*   Sentiment |  
    | **Teams and users** | Such as:
*   Activity
*   Assignment behavior
*   Agent performance
*   Capacity |  
    | **Work items and sessions** | Such as:
*   Agent availability
*   Queue management
*   Routing efficiency |  
    | **Custom objects (KObjects)** | Configured business objects such as:
*   Cases
*   Tickets
*   Products |  
    | **Notes and collaboration** | Such as:
*   Internal documentation
*   Collaboration signals |  
    | **Time-based metrics** | Such as:
*   Business hours vs. calendar time
*   Queue time |

Common insight types produced from these sources include:

*   Comparative analysis (Correlating CSAT scores withvariables like channel, team, response time, tags, and more)
*   Historic trends and seasonal patterns
*   Productivity distribution
*   Queue efficiency diagnostics
*   Response time trends
*   Revenue attribution (when configured)
*   Sentiment patterns
*   Staffing peak identification
*   Time-between-contact analysis

**Note:** Use the Help widget for greater transparency into Data Explorer's capabilities. Data Explorer includes helpful prompting guidelines, data access and analysis capabilities, visualization support, and more. The insights provided are specific to your Kustomer environment and are based on your data.

### Prompting best practices

High signal prompts are specific about time, dimensions, and business goals. Use this structure:

**\[Time window\] + \[Metric\] + \[Breakdowns\] + \[Filters\] + \[Business goal\]**

Examples of high signal context to include:

*   Clear time periods. Ex: last 30 days, last full quarter, same week last year
*   Relevant dimensions. Ex: channel, team, agent, customer segment, tag, queue
*   Outcome or goal. Ex: capacity planning, triage policy change, coaching plan
*   Follow-ups. Ask for comparisons or trend explanations to refine insights

#### Anti-patterns to avoid

*   Vague time frames such as "recently" when you need month-over-month analysis
*   Missing dimensions when you plan to compare performance across teams or channels
*   Asking for unsupported external data sources within the same answer

### Troubleshooting and FAQs

**I do not see Data Explorer in Reporting**

*   Verify AI-enabled features are turned on for your organization
*   Confirm you have access to Reporting. Ask your admin to review your permissions

**My results look off**

*   Add a clear time window and specify the dimensions you expect
*   Ask "show me the filters and definitions you used" to validate scope
*   Follow with "compare to last period" to spot regression or seasonality

**Can I ask about my custom objects?**

*   Yes. Reference the object by the name you configured and the attributes you want to analyze

**What time zone is used?**

*   Data Explorer respects your workspace time configuration when analyzing time-based metrics

**Can I use the results in Custom reports?**

*   [Reporting assistant is available in Custom Reports](https://kustomer.kustomer.help/en_us/custom-reports-HkewhVHIW).

#### Team Pulse vs. Data Explorer

*   **Team Pulse** displays real-time agent activity, including active work-items. Work-items are only visible in Team Pulse under individual agent details and cannot be surfaced through a custom search.
*   **Data Explorer** pulls data from the backend and can answer questions about historical conversation counts, resolution times, and agent activity. Data Explorer results may differ from Team Pulse because they query different data sources.

Use Team Pulse for real-time monitoring and Data Explorer for historical analysis.

### Upcoming enhancements

We're continuously expanding the Data Explorer's capabilities to make it even more powerful and useful for your team. Here are some exciting features we're working on:

*   Use Data Explorer analysis in Custom reports.
*   Use Data Explorer analysis in Standard reports.
*   Use Data Explorer to create Custom report charts or entire Custom reports.
*   Get Data Explorer analysis to surface business risks you face now before they grow into long term business challenges.