Dashboard filters

Add interactive filters to dashboards to let users explore data across multiple questions without creating duplicate dashboards.

Dashboard filters allow people to change what data is displayed or how it’s displayed across multiple questions at once. Instead of creating separate dashboards for different time periods or categories, you can create one flexible dashboard with filters.

Filters vs parameters

There are two types of widgets you can add to dashboards:

Filters

Filters determine WHAT data to show

Examples:

• Show only orders from the last 30 days
• Display data for a specific region or category
• Filter to customers with high lifetime value

Filter types:

• Date picker
• Location
• Text or category
• Number
• ID
• Boolean

Parameters

Parameters determine HOW to show data

Examples:

• Change time grouping from daily to monthly
• Switch aggregation method
• Adjust visualization granularity

Parameter types:

• Time grouping

Adding a filter to your dashboard

1. Enter edit mode

   Click the pencil icon to start editing your dashboard.

2. Choose filter location

   You can add filters to:

   • The dashboard itself (top bar): Visible across all tabs
   • Heading cards: Scoped to cards below the heading on current tab
   • Question cards: Only affects that specific card

   Click the filter icon where you want to add the filter.

3. Select filter type

   Choose the type that matches your data:

   • Date picker (for dates)
   • Text or category (for labels and categories)
   • Number (for numeric values)
   • Location (for geographic data)
   • ID (for IDs and codes)
   • Boolean (for true/false values)
4. Configure and connect

   Set up the filter options and connect it to the cards you want it to affect.

5. Save your dashboard

   Click Save to apply your changes.

Filters only appear on tabs where they’re connected to at least one card. If a filter isn’t connected to any cards on the current tab, it won’t be visible.

Date picker filters

Date filters let users filter time-based data:

Single date

Pick one specific date using a calendar widget.

Best for:

• Viewing data for a specific day
• Selecting an exact date point

Date range

Select start and end dates to define a period.

Best for:

• Custom time periods
• Comparing specific date ranges

Relative date

Dynamic date ranges that update automatically.

Examples:

• Previous 7 days
• Current month
• Last quarter
• Next 30 days

Best for:

• Always-current dashboards
• Rolling time windows

Month and year

Select a specific month and year from dropdowns.

Best for:

• Monthly reports
• Year-over-year comparisons

Quarter and year

Pick a quarter (Q1, Q2, Q3, Q4) and year.

Best for:

• Quarterly business reviews
• Seasonal analysis

All options

Provides all date filtering options in one widget - just like the query builder’s date filter.

Use relative dates for dashboards that should stay current. “Previous 30 days” automatically adjusts to always show the last 30 days, no matter when you view the dashboard.

Text or category filters

Filter based on text values, labels, or categories:

Filter operators

• Is: Matches specific values (dropdown or search)
• Is not: Excludes specific values
• Contains: Matches values containing the text
• Does not contain: Excludes values with the text
• Starts with: Matches values beginning with the text
• Ends with: Matches values ending with the text

Widget types

Dropdown list

Shows all possible values in a selectable list with checkboxes.

Pros:

• Loads instantly from cache
• See all options at once
• Easy to select multiple values

Best for:

• Columns with fewer than 100 unique values
• When users need to see all options
• Frequently filtered fields

Search box

Type to search for values with auto-suggest.

Pros:

• Works well with many values
• Fast for users who know what they want
• Supports multiple selections

Best for:

• Columns with 100-1000 values
• When users know what they’re looking for
• Lists of names, products, or IDs

Input box

Plain text input without suggestions.

Pros:

• Simple and fast
• Works with any number of values
• Good for partial matching

Best for:

• Columns with many unique values
• Free text fields (comments, descriptions)
• Using “contains” or “starts with” operators

If you don’t see the dropdown list option, an admin needs to enable it in the metadata settings for that column, or you need to map the model column to a database field.

Number filters

Filter based on numeric values:

• Equal to: Exact match
• Not equal to: Exclude specific value
• Between: Range of values
• Greater than or equal to: Minimum threshold
• Less than or equal to: Maximum threshold

Example uses:

• Orders over $1,000
• Products with inventory between 10 and 50
• Accounts with exactly 5 users

Location filters

Filter geographic data:

Country

Select one or more countries from a searchable list.

State

Filter by state or province.

City

Select specific cities.

ZIP / postal code

Filter by ZIP or postal codes.

All location filters support the same operators as text filters (is, is not, contains, etc.).

ID filters

Simple input for IDs, order numbers, or unique identifiers:

• Can accept single or multiple values
• Plain text input box
• Useful for looking up specific records

Example: Filter dashboard to show data for customer IDs 1001, 1002, and 1003.

Boolean filters

Filter true/false or yes/no values:

• Simple toggle or dropdown
• Good for status flags (active/inactive, paid/unpaid, etc.)
• Can also handle null values (empty/not empty)

Time grouping parameters

Change how time-based data is grouped across the dashboard:

1. Add time grouping parameter

   Select Time grouping when adding a new filter.

2. Connect to date fields

   Map the parameter to datetime columns in your questions.

3. Users can change grouping

   Choose to view data by:

   • Minute / hour
   • Day / week / month
   • Quarter / year
   • Day of week, month of year, etc.

Time grouping parameters don’t filter data - they change the aggregation level. All data points remain; they’re just grouped differently.

For query builder questions, time grouping parameters can only connect to date fields in the final stage of the query. If your last stage is a filter or sort, time grouping won’t work on earlier date fields.

Connecting filters to cards

After adding a filter, connect it to the questions you want it to affect:

1. Click on a card's dropdown

   While editing the filter, each card shows a dropdown to connect it to the filter.

2. Select the field to filter

   Choose which column in that question the filter should affect.

3. Repeat for other cards

   Connect the filter to all relevant cards on the dashboard.

Auto-connecting

When you connect a filter to a card by selecting a field, Metabase offers to automatically connect the filter to other cards that have the same field. This saves time when you have many cards with the same column.

Metabase will also try to auto-connect new cards you add later if they contain the field.

To undo auto-connecting, click the notification that appears or manually disconnect cards by clicking the X next to the connected field.

Connecting filters to SQL questions

For SQL questions to work with dashboard filters, the question must include variables:

SELECT *
FROM orders
WHERE {{status}} AND {{order_date}}

Then:

1. Add the SQL question to your dashboard
2. Create a dashboard filter matching the variable type
3. Connect the dashboard filter to the SQL variable

Use field filters in SQL questions (not basic variables) for better compatibility with dashboard filters and more flexible filter options.

Connecting filters to text cards

Text cards can use filter values to create dynamic content:

# Performance report [[for {{region}}]]

Showing data [[from {{start_date}} to {{end_date}}]]

Variables:

• Use {{variable}} to insert the filter value
• Wrap text in [[ ]] to show it only when the filter has a value

Filter configuration options

Setting a default value

Provide a starting value when the dashboard loads:

1. Click the filter while editing
2. Set the Default value in the sidebar
3. Users can change it, but it starts with your default

Default values help users understand what the filter does and provide a good starting point for exploration.

Requiring a filter

Force users to select a filter value:

1. Click the filter while editing
2. Toggle Required in the sidebar
3. Set a default value (required when filter is required)

Use required filters when:

• Unfiltered queries would return too much data
• The dashboard only makes sense with a filter applied
• You want to reduce database load by preventing broad queries

Multi-select filters

Allow selecting multiple values:

1. Click the filter while editing
2. Under People can pick, select “Multiple values”
3. Save the dashboard

Dropdown and search box widgets show checkboxes for multi-select.

Custom selectable values

Control what values appear in dropdowns:

From connected fields

Default behavior - shows all values from the connected columns.

From another model or question

Pull filter values from a different data source:

1. Select this option
2. Choose a model or saved question
3. Pick which column contains the filter values

Useful for:

• Curated lists of values
• Active/current records only
• Custom sort orders

Custom list

Manually enter specific values:

1. Select this option
2. Type each value (press Enter after each)
3. Reorder by dragging

Best for:

• Small, fixed lists
• Custom groupings
• Specific value sets

Filter organization

Reordering filters

1. Enter dashboard edit mode
2. Click the grabber handle (six dots) on the left of a filter
3. Drag to reposition

Renaming filters

1. Enter dashboard edit mode
2. Click the filter
3. Change the label text in the sidebar

Renaming a filter only changes the display label. It doesn’t affect which fields the filter is connected to.

Removing filters

1. Enter dashboard edit mode
2. Click the filter
3. Click Remove in the sidebar
4. Save the dashboard (or click Cancel to undo)

Linked filters

Create filter dependencies where one filter’s options depend on another filter’s value.

For example:

• Select a country first, then state filter only shows states in that country
• Choose a category, then product filter shows only products in that category

This creates intuitive, cascading filter experiences.

See the Linked filters documentation for detailed setup instructions.

Auto-apply filters

Control when filters refresh the dashboard:

Auto-apply ON (default)

Dashboard refreshes immediately when you change any filter value.

Best for:

• Fast-loading dashboards
• Single-filter interactions
• Immediate feedback preference

Auto-apply OFF

Dashboard refreshes only when you click the Apply button.

Best for:

• Slow or complex dashboards
• Adjusting multiple filters before refreshing
• Reducing database load
• Controlling when queries run

To toggle auto-apply:

1. View the dashboard (not in edit mode)
2. Click the three-dot menu (…)
3. Select Edit settings
4. Toggle Auto-apply filters

Using dashboard filters

Once filters are set up, users can:

1. Click the filter widget to open the selection interface
2. Choose value(s) from dropdown, search box, or calendar
3. View updated results (or click Apply if auto-apply is off)
4. Clear the filter by clicking the blue X

Filter selections are personal to each user’s view. If someone else views the same dashboard at the same time, they won’t see your filter selections.

Best practices

Start with key filters

Place the most commonly used filters first. Users scan left-to-right.

Use clear labels

Name filters descriptively: “Order Status” not just “Status”, “Customer Region” not “Region”.

Set helpful defaults

Provide default values that show meaningful data immediately.

Choose the right widget

Use dropdowns for short lists, search boxes for longer lists, input boxes for free text.

Group related filters

Keep filters for similar concepts together (all location filters, all time filters).

Don't over-filter

Too many filters overwhelm users. Aim for 3-5 filters; use tabs for different views instead.

Dashboard filter permissions

Users can interact with dashboard filters based on their permissions:

• View access: Can use filters and see filtered results
• Curate access: Can add, edit, and remove filters

Data access permissions still apply - filters can’t show data users don’t have permission to see.

Next steps

Creating dashboards

Learn more about building effective dashboards

Subscriptions

Send filtered dashboard results via email or Slack

SQL editor

Add variables to SQL questions for dashboard compatibility