Plans and permissions:
- Add-on: Copilot, billed separately
- View the Copilot impact report: Admin or Manager role, or a custom role with Copilot impact set to Can view
- A role can be limited to certain teams or queues in this report.
Use the Copilot impact report to see how Copilot is helping your team. It connects each support team member’s Copilot usage to their performance, such as handle time, productivity, and CSAT, and shows which Copilot features your team uses most. Use it to track adoption, find coaching opportunities, and show the value of Copilot.
To open it, go to Report > Copilot impact.
How the report works
Copilot is Assembled’s AI assistant for your support team: it drafts replies, answers questions, summarizes and wraps up cases, and translates. The report groups support team members by how much they use it. Each week, everyone is placed in a usage level based on the share of their assigned cases where they used a Copilot feature:
| Usage level | Support team members in this group |
|---|---|
| High (> 75% of cases) | Used a Copilot feature on more than 75% of their cases. |
| Moderate (25–75%) | Used a Copilot feature on 25–75% of their cases. |
| Low (1–24%) | Used a Copilot feature on 1–24% of their cases. |
| None (0%) | Didn’t use any Copilot features on their cases. |
| No Copilot access | Don’t have Copilot features turned on. |
Usage levels are updated weekly, so people can move between them as their usage changes.
To narrow the report, choose a date range and filters: Assignee availability, Permissions, Team, Queue, Site, and Channel. To compare groups, group the data by channel, team, queue, or site.
Overview tab
The Overview tab shows your team’s performance through the lens of Copilot usage.
- Weekly Copilot usage rate shows how your team is adopting Copilot, week over week, by usage level.
- Agent performance metrics shows how each usage level performs over time, based on last week’s usage. Choose the metric at the top right of the chart.
- The table below the charts shows the same performance data for individual support team members, or for groups such as sites, teams, queues, or channels. Sort or group it to dig deeper.
Depending on your company settings, you can choose these metrics:
| Metric | What it measures |
|---|---|
| Avg handle time | The average time each person spends handling solved cases. It shows how Copilot use relates to the time it takes to solve a case. |
| Solved (per occupied hour) | How many cases are solved per hour that people are logged in and working on customer issues. It shows how Copilot use relates to productivity. |
| CSAT | Customer satisfaction scores from your connected customer service platform. It shows how Copilot use relates to customer satisfaction. See How CSAT is calculated. |
Features tab
The Features tab breaks down Copilot usage by what support team members do, such as drafting replies, wrapping up cases, and rephrasing messages.
- Total feature usage shows how often your team uses each feature, so you can see which are most helpful and which are underused.
- Copilot usage breakdown shows how often each person uses each feature. Use it to monitor adoption and find coaching opportunities.
What the report doesn’t show
The Copilot impact report shows how Copilot helps your support team. For other data, use these reports:
| Data | Report |
|---|---|
| Your team’s other performance metrics, such as adherence or scheduled utilization | Report > Team performance |
| How AI agents help your customers in chat, email, voice, and other channels | Report > AI agent performance |
| Your overall case volume and service level | Report > Support volume |
Share the report
To share the report with someone who has an Assembled account, send them the page link. They need to sign in, and their role needs access to the report. See Roles and permissions.
To share the data outside Assembled:
- Open the tab you want to share.
- Set the date range and filters.
- Select Download.
The data downloads as a CSV file. To remove sensitive performance data before you share it, edit the file.
How CSAT is calculated
For Zendesk, Assembled reads each ticket’s satisfaction rating, which is one of these:
| Rating | What it means |
|---|---|
| Unoffered | The customer wasn’t asked to rate the interaction. |
| Offered | The customer was asked, but didn’t rate it. |
| Good | The customer rated the interaction as good. |
| Bad | The customer rated the interaction as bad. |
Only Good and Bad ratings count toward CSAT:
CSAT = Good ratings ÷ (Good ratings + Bad ratings)
For example: a support team member receives 8 good ratings and 2 bad ratings. Their CSAT is 8 ÷ (8 + 2) = 80%. Tickets that were unoffered, or offered but not rated, don’t change it.
Where Assembled splits CSAT by Copilot use, CSAT with Copilot counts the tickets Copilot was involved in, and CSAT without Copilot counts the rest, using the same calculation.
When CSAT data appears
Assembled starts collecting Copilot data as soon as your support team starts using it, and calculates CSAT once a day. Expect to see new data the next day.