Copilot brings AI to your service workflows, helping agents respond faster and assisting customers sooner. To understand Copilot’s value, you need reliable, actionable data. Before getting started, establish a clear baseline; knowing where you started makes it easier to measure improvements and prove ROI over time.
Zendesk analytics help you monitor how Copilot tools impact ticket handling, agent performance, and customer satisfaction. This article walks through best practices on where to find prebuilt dashboards, which metrics to monitor, and how to customize and share your reporting.
This article contains the following topics:
Using prebuilt dashboards to measure performance
The following prebuilt dashboards can help you to monitor and measure your Copilot performance:
Zendesk Copilot: Agent productivity dashboard
The Zendesk Copilot: Agent productivity dashboard is your command center for monitoring how agents interact with Copilot AI in Zendesk. The dashboard is organized into two tabs: AI suggestions and AI agent tools. It offers a clear view of adoption and engagement across key AI tools, helping you track usage, optimize workflows, and drive productivity.
Here's an overview of the two tabs:
- AI suggestions: Measures engagement and performance of Copilot features like similar tickets, merging suggestions, suggested macros, suggested replies and quick answers.
-
AI agent tools: Captures how agents use writing tools such
as expand, make formal, make friendly and ticket summary (we plan to expand
this tab to include enhanced writing tools like
simplify, custom prompt, and rewrite in your tone, as well as new
dimensional filters, such as channel, ticket ID and others).
This tab mirrors the current Generative AI Agent Tools dashboard. Because it includes the same insights, we will deprecate the existing Generative AI Agent Tools dashboard in Q3 2025 as part of the Dashboard Builder transition process.
Key things to look for:
- Compare tickets solved with AI to those without and measure high-level agent adoption. Then look at specific agents’ AI usage and cross-reference with their overall performance.
- Spot agents with high AI suggestion acceptance rates. Use this data to recognize and reward early adopters, or involve them in peer-to-peer enablement and best practice sharing.
- Analyze how features like suggested replies, merging suggestions, and ticket summaries are being used. If adoption is low, investigate why, then update training or processes. If adoption is high, scale those behaviors across your team to improve resolution speed and consistency.
- Track how AI tools are impacting key metrics like agent productivity, resolution time, and customer satisfaction. By correlating AI adoption with improved support outcomes, you can clearly demonstrate the value of your AI investment to leadership.
- Use the Zendesk Copilot: Agent productivity dashboard to compare wait times for tickets with and without generative AI. For deeper insight into agent handling time, we recommend the Time Tracking app from the Zendesk Marketplace. This app helps you measure how long agents spend on each ticket. To get meaningful results, make sure you're comparing similar ticket types. For example, simple tickets or those automatically solved are typically faster to resolve and may not require AI assistance, which can skew the data if not properly segmented.
Intelligent triage dashboard
The Intelligent triage dashboard provides valuable insights into your tickets, which you can use to improve your workflow and determine how effective your service is. You can view the most common ticket intents, check sentiment trends, and review how well triage is distributing work. The intelligent triage dashboard has four tabs with detailed views of intent, language, and sentiment.
Key things to look for:
- Identify the top intents of incoming requests and monitor KPIs across AI insights, allowing teams to make key optimization decisions.
- Review one touch tickets by top intents and use those insights to automate processes.
- Reduce escalations by identifying low-sentiment intents. Use customer sentiment reporting to spot common topics with negative sentiment, then build or refine escalation workflows to proactively address them before they escalate.
- Optimize resource allocation across teams by reviewing ticket distribution across different languages.
- Look for intents or languages where resolution takes longer or requires multiple replies. These patterns can point to training or workflow improvement opportunities.
Building custom reports with key Copilot metrics
If you want more control over what you track, Zendesk analytics lets you build custom reports using relevant Support and Copilot data. Here are some of the common metrics customers tend to measure before and after they implement copilot:
| Metric | What is it | Why it matters | Best practices / Actions |
| First response time | Time between ticket creation and the first agent response. | Shows how quickly your team is responding to incoming customer requests. | Compare difference between first response time of tickets using Copilot to those that didn’t. |
| CSAT | Customer satisfaction rating after a ticket is resolved. | Helps you understand how Copilot tools are impacting customer experience. | Segment by AI-assisted vs non-AI tickets to assess where satisfaction improves or drops. Use this to refine AI deployment. |
| Average resolution time | Time from ticket creation to final resolution. | Tracks efficiency across the ticket lifecycle. | Compare AI vs non-AI tickets to see where Copilot reduces resolution time. Review by intent to optimize processes. |
| Automated resolution rate | Percentage of tickets resolved without a human agent. | Measures how much value AI is delivering independently. | Monitor trends over time. Look for intents with high automation potential to expand coverage. |
| First contact resolution rate | Tickets resolved with a single agent reply. | Indicates the effectiveness of Copilot-generated responses. | Filter by AI usage to identify which intents and agents benefit most. Use insights to improve content or training. |
| NPS | Measures customer loyalty over time. | Gives a big-picture view of how AI influences long-term satisfaction. | Compare before and after Copilot implementation to track overall sentiment shift. Look for patterns in score changes. |
Sharing results with stakeholders
After creating dashboards and reports, consider sharing them with key stakeholders to keep teams aligned and show the impact of Copilot over time. You can schedule dashboards to be sent out on a one time or regular cadence. This is useful for sharing updates monthly, quarterly, or after deploying Copilot at your organization. Dashboards can also be exported to share in presentations or leadership updates.
Helpful resources
Review the following articles to learn more about measuring the impact of Copilot with analytics:
- Reporting on the auto assist feature of agent copilot
- Zendesk AI insights & reporting
- Zendesk analytics recipes reference
- Working with prebuilt dashboards
- Explore recipe: Intelligent triage predictions and confidence
- Explore recipe: Intelligent triage changes to intent
- Explore recipe: Comparing intelligent triage intent predictions with a custom About field
- Explore recipe: Breakdown of CSAT by customer sentiment