Source: https://atomicwork.docs.pageloop.ai/product/analytics/executive-views/insights-report

# Insights report

The **Insights report** helps [admins and analysts](/product/identity-access-management/people-roles-and-teams/roles-in-atomicwork) make sense of what’s happening across their Atomicwork platform — from how employees are using Atom to how efficiently their teams are resolving support requests.

Conversations and requests often reflect two sides of the same problem. But without a connected view, it’s hard to know where automation is working, where requests are piling up, or why SLAs are being missed. This report brings both together into a single, theme-based view. You can see which types of employee conversations are leading to requests, where Atom is helping deflect them, and where manual effort is still required — giving you the clarity to take targeted action and improve outcomes across the board.

To view the Insights report, go to **Analytics > Atom > Insights** in the left navigation bar.

- **Org admins and analysts** can view insights across all workspaces.

- **Workspace admins and analysts** see only the workspaces they have access to.

## Understanding the Insights report

The report introduces the **top 10 themes in your platform or workspace**, ranked by volume. Each theme groups together related conversations and requests to give a clear picture of where users are seeking help or experiencing friction.

Themes help admins:

- Identify what topics are consuming the most attention.

- Understand where automation is succeeding or failing.

- Uncover process bottlenecks across departments.

Clicking a theme expands a list of **sub-themes**, which are more specific clusters of issues nested under a broader topic. For example, a theme like _Okta issues_ might include sub-themes like _MFA resets_ and _Account lockouts_.

Against each theme and sub-theme, you’ll see:

- **Total conversations**: The number of unique conversations related to that topic.

- **Requests created**: The total number of requests associated with the theme  created via Atom (on Slack/Teams or the browser extension), the portal, or email.

### Exploring sub-themes

Clicking into a sub-theme provides a detailed breakdown of that specific issue. This view helps you get a clearer sense of how that issue is unfolding across channels. It’s where you can break down Atom’s impact, request load, and automation gaps, scoped down to a specific problem area.

#### Filtering by duration or workspace

The report has two filters that allow you to tailor the data you're viewing:

- **Duration**: Specifies the time window for metrics displayed. You can choose from presets like _This week_, _This month_, _Last month_, or set a custom range.

- **Workspace**: Focusses insights on a specific workspace (like IT, HR, Marketing, etc.) or across the entire platform when \_All workspaces \_is selected.

#### Understanding the metrics

Each sub-theme includes a snapshot of key metrics to help you quickly assess volume and impact.

For each metric, the chart displays two values:

- One for the specific day, week, or month such as _Conversations deflected_

- One for the selected duration such as\_ Total conversations deflected\_

This helps you spot trends and outliers while keeping the broader context of overall performance in view.

| **Metric**              | **Description**                                                                              | **Why it matters**                                                               |
| ----------------------- | -------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------- |
| Conversations           | Number of unique conversations on this issues                                                | Measures how often employees engage with Atom about this issue                   |
| Conversations deflected | Conversations handled completed by Atom without human intervention or a request being raised | Gauges automation success; a high number suggests Atom is resolving queries well |
| Requests created        | Total number of requests tied to this topic across all channels                              | Indicates real demand—includes tickets via Atom, portal, or email                |

These metrics allow you to identify:

- Topics where Atom is reducing load effectively (high deflection)

- Themes that are creating too many requests (high request volume, low deflection)

- Themes that are surfacing heavily through direct channels like portal or email (requests without conversations)

For example, a sub-theme like _Monitor not turning on_ might show high conversation volume, high volume of conversations deflected and very few requests. This suggests that Atom is successfully resolving most queries without escalation.

In contrast, a sub-theme like _VPN access_ might have very few conversations deflected but hundreds of requests suggesting a content gap or lack of awareness.

### Conversations view: Atom's performance

Clicking **the Conversations tab** below the overview highlights two important contributors to Atom’s performance within each sub-theme:

- **Top contributing knowledge:** Lists the articles Atom used most frequently to respond to user queries. It includes [internal documentation](/product/ai-knowledge/knowledge-management/add-and-organize-knowledge), [verified answers](/product/ai-knowledge/knowledge-management/verified-answers), and [trusted knowledge sources](/product/ai-knowledge/knowledge-management/trusted-knowledge-for-employee-questions) — reflecting what Atom actually used in conversations, not just what exists in your knowledge base.

- **Top contributing catalogs:** Shows the catalog items Atom displayed most often in related conversations, helping verify whether the right forms are being surfaced.

Let’s say you notice that **_MFA reset_** is the highest volume sub-theme under _Okta issues_. In the snapshot, you see:

- Conversations: 80

- Deflected conversations: 5 (deflection rate is \~6%)

- Requests created: 100

- No contributing knowledge, but the default incident form is listed under Top contributing catalogs

This tells you that:

- Atom isn't equipped to resolve MFA reset questions because it has no useful articles or service items to pull from

- **No contributing knowledge** suggests there’s a content gap — employees aren’t being shown relevant guidance. **What you should do:**

1. **Add a targeted knowledge article:** Create a doc or a verified answer that clearly explains how to reset MFA in Okta

2. **Create a dedicated catalog item:** Build a short, purpose-built incident form specifically for MFA resets that includes fields device type, last successful login, etc.

3. **Improve Atom's mapping:** Use clear descriptions and tags like MFA reset, Okta, etc. to ensure Atom surfaces the new article/catalog item.

4. **Track changes:** Monitor whether conversations increase, deflection improves and request volume decreases or shifts toward the new catalog item

### Requests view: Operational performance

Switching to the **Requests** tab reveals metrics related to request handling like responsiveness, efficiency, and reassignment trends.

| **Metric**                 | **What it measures**                                                    | **Why it matters**                                             |
| -------------------------- | ----------------------------------------------------------------------- | -------------------------------------------------------------- |
| **Time-based metrics**     |                                                                         |                                                                |
| Time to first assignment   | Time from request creation to first assignment                          | Shows how quickly requests are being picked up for action      |
| Time to first response     | Time from request creation to first agent response                      | Indicates responsiveness to employees                          |
| Time to resolution         | Total time from request creation to closure                             | Measures overall resolution efficiency                         |
| **Reassignment metrics**   |                                                                         |                                                                |
| Reassigned requests        | Percentage of resolved requests that involved at least one reassignment | HHigh numbers may point to unclear ownership or routing issues |
| Median reassignments       | Median number of handoffs per resolved request                          | Highlights friction in internal request handling               |
| **SLA breach rates**       |                                                                         |                                                                |
| First response breach rate | Percentage of requests that missed the SLA for first response           | Reflects how well teams meet response expectations             |
| Resolution breach rate     | Percentage of requests that missed the SLA for resolution               | Signals overall service quality and turnaround times           |

Insights is designed to be the central lens through which admins and analysts measure both Atom performance and service management workflows. With a focus on theme-level analytics and actionable breakdowns, you’re not just tracking numbers — you're uncovering opportunities to improve employee experience, efficiency, and quality of service.