Getting Started with Custom Metrics in Zendesk Explore

Getting Started with Custom Metrics in Explore

Learn what calculated metrics are, why they matter, and how to build your own in Zendesk Explore using Geckoboard's custom metrics and best practices.

Why you need custom metrics

Out-of-the-box Explore dashboards provide a strong foundation, but CS managers often need more actionable insights. For instance, you may want to measure the percentage of inbound calls that were abandoned in IVR, or highlight tickets resolved in less than 20 minutes. These views are not possible without creating calculated metrics.

By embedding logic directly inside a metric, you can “lock in” the exact conditions you want. This avoids conflicts with dashboard filters and ensures results are consistent across reports. It also allows you to place different subsets of data side by side for clear comparison.

The core structure

Most calculated metrics use the IF/THEN/ENDIF pattern. This logical structure lets you tell Explore: if these conditions are true, then count or sum this value.

Creating Your Own Metric with Hard-Coded Filter demonstrates this clearly. It shows how to embed filter logic inside the metric itself, rather than relying on a dashboard filter. This is particularly useful when comparing subsets of data that would otherwise overlap.

Examples from the library

Here are some calculated metrics from the library that illustrate what is possible:

Short-Abandoned Calls (Metric)
Counts inbound calls abandoned in the queue within 10 seconds. Helps exclude instant hang-ups from your main abandonment KPI while still tracking them separately.

Abandoned in IVR Calls
Captures inbound calls terminated while navigating the IVR menu. High values here can point to confusing menu design.

Abandoned in On-Hold Calls
Counts calls where customers disconnected during hold time. This highlights potential issues with agent capacity or hold messaging.

Abandoned Inbound Calls %
Shows the headline abandonment rate as a percentage of inbound calls, useful alongside the more granular metrics above.

Long Abandon Rate %
Focuses on abandons after a defined wait threshold to give a fairer view of demand agents could have served.

Tickets Solved in Less than 20 Minutes
Flags tickets resolved within 20 minutes of business time. Useful for measuring efficiency and for setting as a team-wide target (the number can easily be changed).

True One-Touch Talk Tickets
Identifies Talk tickets resolved entirely within a call, with no follow-up reply. Essential for measuring efficiency on the Voice channel.

Non-answered Inbound Calls %
Demonstrates that calculated metrics can also be straightforward arithmetic, combining built-in measures to create percentages, ratios, or other derived values:

D_COUNT(Non-answered inbound calls) / D_COUNT(Inbound calls)

This shows that calculated metrics are not limited to condition-based logic. They can also use simple division or other arithmetic to provide a different perspective on your data.

Building your first custom metric

Follow these steps to create a calculated metric using code:

Go to Explore → Report Builder and select the dataset you want to use, such as Support - Tickets or Voice - Calls.
Click the Calculator icon in the right-hand panel.
Choose Standard calculated metric.
Copy the code from the library and paste it in. Replace placeholders such as [Attribute] or "Value" with the exact field names and conditions you want. A great tip here is to use the Select a Field button to ensure you get the name correct. It also helps you find fields when you’re not quite sure of the name.
Check that you have the green validation tick to ensure your code is correct. Best practice is to start by pasting in the code, check that you have the green tick, then change just one section at a time to narrow down which part might be causing you trouble.
Save the metric with a clear name, for example “Tickets solved in less than 20 minutes”.
Under Metrics on the left, select Add, and find your new Calculated Metric under the aforenamed section, or you can search for it.
Add an attribute such as Agent Name, Year and Week, or something else to view your metric in helpful charts.

Best practice tips

Use the code you find in this library as templates to build from—they provide working formulas you can adapt.
Document your work. It’s recommended to keep a central record of custom metrics and their purpose.
Beware dependencies. Some metrics rely on others being built first, such as % True One-Touch Talk Tickets needing True One-Touch Talk Tickets to work.
Always validate your results carefully. Try testing results against sample tickets or calls.

Summary

Calculated metrics in Explore let you go beyond default reporting. They can use conditional logic, as in Short-Abandoned Calls (Metric) or Abandoned in IVR Calls, or simple arithmetic, as in Non-answered Inbound Calls %. Together, these approaches unlock a wide range of insights tailored to your business.

Once you are familiar with the IF/THEN/ENDIF structure and simple formulas, creating metrics with code becomes a straightforward process that adds real value to your reporting.