Create a custom metric in Analytics

A custom metric turns the data in a published model into a reusable calculation that you can add to dashboard charts.

Custom metrics are available through the Early Access Program for sites on Premium and Enterprise plans. Your site must be enrolled before you can create them. To join, submit a request at the Early Access Program sign-up form.

Before you begin

• Create and publish a data model in Analytics.

• Make sure the model contains every column you want to calculate, group, or filter by.

• Make sure you have access to the model.

To open Metrics, select App switcher, then Atlassian Analytics, then Metrics in the sidebar.

Select your site and data model

A custom metric belongs to one site and can only use models from that site. Use the site picker to select the site that contains your published model.

Select the model you want the metric to use. A model must be published before you can build a metric on it.

Select how to calculate the metric

For an aggregated metric, select Count of unique, Sum, Average, Min, or Max, then select the column to aggregate. For example, you can count unique work item keys to measure the number of incidents linked to applications, or average a formula column to measure change implementation lag.

If you want to return individual detail rows instead of calculating a value, create an unaggregated metric. Unaggregated metrics appear as tables when you add them to dashboards.

Preview the results

Use the preview to check that the metric returns the data you expect. You can group the preview by a column from the data model, such as application name, application tier, or work item assignee.

The preview helps you validate the metric's data. You select the final visualization when you add the metric to a dashboard.

Filter your metric

Add a metric filter when every chart built from the metric should use the same subset of data. Any filter applied to the metric also applies to its charts.

Filters inherited from the data model continue to apply as well.

Name and publish your metric

Give the metric a name that describes what it measures, then publish it. A metric only becomes available in dashboards after you publish it.

One data model can support multiple metrics. For example, the same deployed-changes model can support an average change implementation lag metric and an unaggregated metric that lists the individual changes.

Control who can edit your metric

Anyone who can access the underlying model can see the metric. Use the metric's share settings to control who can edit it.

Model access and existing data permissions continue to apply. If someone can't access the model, they won't see the metric or any chart built from it. People who can access the model only see data they already have permission to view.

Where you can use a custom metric

Published custom metrics can appear in Jira Service Management Summary dashboards, Assets dashboards, and Platform Dashboards in Open Beta for the same site.

Limitations

During the Early Access Program:

• A metric can only use published models from one site.

• An unaggregated metric can only be displayed as a table and appears under the List widget in dashboards.

• You can create up to 1,000 metrics.