Metrics

Define official calculations for important business numbers

Metrics are pre-defined calculations that standardize how your team measures key business values. Create your aggregations once, save them as metrics, and use them consistently across all questions.

Metrics ensure everyone calculates important numbers like revenue, conversion rate, or active users the same way.

Why use metrics

Metrics solve the problem of inconsistent calculations across your organization. Instead of five different people calculating revenue five different ways, you define it once as a metric that everyone can reference.

Where you can use metrics

Collections

Save metrics to collections for easy organization and access

Dashboards

Add metrics directly to dashboards for at-a-glance insights

Database browser

View available metrics when browsing your databases

Query builder

Select metrics as aggregation columns in the Summarize block

Using metrics in questions

Metrics integrate seamlessly into the query builder workflow.

Common metrics in summarization

When building questions, find your team’s metrics in the summarization section under Common metrics. You can:

• Select multiple metrics at once
• Combine metrics with different dimensions
• Use custom expressions to perform calculations on metrics

Metrics as data sources

You can start a question directly from a metric. If the metric includes a time dimension, Metabase automatically groups by that dimension. You can change the grouping to analyze the metric by other dimensions.

Use custom expressions to combine multiple metrics or perform advanced calculations like growth rates or ratios.

Creating a metric

1. Open the metric creation dialog

   Use the command palette (Cmd/Ctrl + K) and search for “New metric”, or navigate to Browse > Metrics in the sidebar and click +

2. Select your starting data

   Choose a model, existing metric, table, or saved question as your data source

3. Define your formula

   Create the aggregation that calculates your metric value. This is the required core of your metric

4. Add optional settings

   Configure joins, filters, and a default time dimension as needed

5. Save your metric

   Give your metric a clear name and description, then save it to a collection

Metric components

The metric editor resembles the regular query builder with key differences:

Formula (required)

The formula is the heart of your metric—the aggregation you’re calculating. You can build metrics on top of other metrics, creating layers of reusable calculations.

Examples:

• Count of orders
• Sum of revenue
• Average order value
• (Total revenue - Total costs) / Total revenue

Default time dimension (optional)

Set a default time dimension to control how the metric displays in collections and dashboards. Without a time dimension, metrics show as single numbers. With one, they display as line charts.

The default time dimension doesn’t lock the metric. Users can still group by different time dimensions and granularities when using the metric in questions.

Joins and filters (optional)

Add joins to bring in data from related tables and filters to focus the metric on specific subsets of data before calculation

Building metrics on metrics

You can create sophisticated calculations by building metrics that reference other metrics:

Metric: Conversion Rate
Formula: [Completed Orders] / [Total Orders] * 100

Where:
- [Completed Orders] is a metric
- [Total Orders] is a metric

This approach creates a single source of truth for each component calculation.

Editing metrics

1. Open the metric

   Navigate to the metric you want to edit

2. Access edit mode

   Click the three-dot menu (…) and select Edit metric definition

3. Make your changes

   Modify the formula, filters, or other settings

4. Save changes

   Save your updates. Editing a metric requires curate access to the metric’s collection

When you click a metric, Metabase assumes you’re using it as a starting point for a new question. Changes you make in this mode create new questions—they don’t affect the metric’s definition. Use Edit metric definition to modify the metric itself.

Time dimension behavior

The default time dimension affects how metrics display, but doesn’t restrict how users can analyze them.

Example scenario:

• You create a “Revenue” metric
• Set default time dimension to “Created At” by month
• Metric displays as a monthly line chart in collections

But users can still:

• Group by quarter instead of month
• Group by “Updated At” instead of “Created At”
• Remove time grouping entirely

Avoid naming metrics with specific time granularities (like “Monthly Revenue”) since users can change the grouping. Use generic names like “Revenue” instead.

Pinned metrics display

When you pin metrics to a collection, their display depends on the default time dimension:

Without time dimension

Displays as a single number card

With time dimension

Displays as a line chart over time

Permissions

Metric permissions work like questions, models, and dashboards through collection permissions.

Viewing metrics

A group can view a metric if they have view access to the metric’s collection. This allows them to:

• See the metric in the collection
• Select it when building queries
• Use it in questions (question permissions depend on the question’s collection, not the metric’s)

Editing metrics

Editing the metric’s query definition requires curate access to the metric’s collection.

Collection permissions interact with data permissions. Users need both collection access to see the metric AND data permissions to query the underlying data source.

Verifying metrics

Available on Pro and Enterprise plans

Admins can mark metrics as verified to indicate they’re trusted and ready for widespread use. Verified metrics display a checkmark badge, helping users identify official calculations.

For more information, see Content verification.

Best practices

Use clear, business-focused names

Name metrics after business concepts (“Monthly Recurring Revenue”) rather than technical details (“sum_of_invoice_amount”)

Add comprehensive descriptions

Explain what the metric measures, how it’s calculated, and any important caveats or filters applied

Build on models when possible

Use models as the foundation for metrics to ensure clean, well-structured data

Layer metrics strategically

Create simple building-block metrics first, then build more complex metrics on top of them

Document calculation changes

When editing metrics, clearly document what changed and why to help users understand the impact

Further reading

Models

Learn how to create models as data sources for metrics

Segments

Create reusable filters to focus your metrics on specific data subsets

Collection permissions

Understand how to control access to metrics

Data permissions

Configure access to underlying data sources