Analytics dashboards

Tier: Ultimate Offering: Self-managed
History
  • Introduced in GitLab 15.9 as an experiment feature with a flag named combined_analytics_dashboards. Disabled by default.
  • combined_analytics_dashboards enabled by default in GitLab 16.11.
  • combined_analytics_dashboards removed in GitLab 17.1.

Analytics dashboards help you visualize the collected data. You can use built-in dashboards by GitLab or create your own dashboards with custom visualizations.

Data sources

A data source is a connection to a database or collection of data which can be used by your dashboard filters and visualizations to query and retrieve results.

Analytics dashboards use the following data sources:

You can also add custom visualization data sources.

Built-in dashboards

To help you get started with analytics, GitLab provides built-in dashboards with predefined visualizations. These dashboards are labeled By GitLab. You cannot edit the built-in dashboards, but you can create custom dashboards with a similar style.

Product analytics dashboards

When product analytics is enabled and onboarded, two built-in dashboards are available:

  • Audience displays metrics related to traffic, such as the number of users and sessions.
  • Behavior displays metrics related to user activity, such as the number of page views and events.

Value Stream Management dashboard

Custom dashboards

Use custom dashboards to design and create visualizations for the metrics that are most relevant to your use case. You can create custom dashboards with the dashboard designer.

  • Each project can have an unlimited number of dashboards. The only limitation might be the repository size limit.
  • Each dashboard can reference one or more visualizations.
  • Visualizations are shared across dashboards.

Project maintainers can enforce approval rules on dashboard changes with features such as code owners and approval rules. Your dashboard files are versioned in source control with the rest of a project’s code.

Dashboard designer

History
  • Introduced in GitLab 16.1 with a flag named combined_analytics_dashboards_editor. Disabled by default.
  • Generally available in GitLab 16.6. Feature flag combined_analytics_dashboards_editor removed.

You can use the dashboard designer to:

Visualization designer

History
  • Introduced in GitLab 16.4 with a flag named combined_analytics_visualization_editor. Disabled by default.
  • Generally available in GitLab 16.7. Feature flag combined_analytics_visualization_editor removed.
note
This feature is only compatible with the product analytics data source.

You can use the visualization designer to:

View project dashboards

Prerequisites:

  • You must have at least the Reporter role for the project.

To view a list of dashboards (both built-in and custom) for a project:

  1. On the left sidebar, select Search or go to and find your project.
  2. Select Analyze > Analytics dashboards.
  3. From the list of available dashboards, select the dashboard you want to view.

View group dashboards

Tier: Ultimate Offering: GitLab.com, Self-managed, GitLab Dedicated
History

Prerequisites:

  • You must have at least the Reporter role for the group.

To view a list of dashboards (both built-in and custom) for a group:

  1. On the left sidebar, select Search or go to and find your group.
  2. Select Analyze > Analytics dashboards.
  3. From the list of available dashboards, select the dashboard you want to view.

View the Value Streams Dashboard

History

To view the Value Streams Dashboard as an analytics dashboard for a group:

  1. On the left sidebar, select Search or go to and find your group.
  2. Select Analyze > Analytics dashboards.
  3. From the list of available dashboards, select Value Streams Dashboard.

Change the location of dashboards

You can change the location of your project or group dashboards.

Prerequisites:

  • You must have at least the Maintainer role for the project or group the project belongs to.

Group dashboards

note
Issue 411572 proposes connecting this feature to group-level dashboards.

To change the location of a group’s dashboards:

  1. On the left sidebar, select Search or go to and find the project you want to store your dashboard files in. The project must belong to the group for which you create the dashboards.
  2. On the left sidebar, select Search or go to and find your group.
  3. Select Settings > Analytics.
  4. In the Analytics Dashboards section, select your dashboard files project.
  5. Select Save changes.

Project dashboards

By default custom dashboards are saved to the current project, because dashboards are usually defined in the project where the analytics data is retrieved from. However, you can also have a separate project for dashboards. This setup is recommended if you want to enforce specific access rules to the dashboard definitions or share dashboards across multiple projects.

note
You can share dashboards only between projects that are located in the same group.

To change the location of project dashboards:

  1. On the left sidebar, select Search or go to and find your project, or select Create new () and New project/repository to create the project to store your dashboard files.
  2. On the left sidebar, select Search or go to and find the analytics project.
  3. Select Settings > Analytics.
  4. In the Analytics Dashboards section, select your dashboard files project.
  5. Select Save changes.

Define a dashboard

To define a dashboard:

  1. In .gitlab/analytics/dashboards/, create a directory named like the dashboard.

    Each dashboard should have its own directory.

  2. In the new directory, create a .yaml file with the same name as the directory, for example .gitlab/analytics/dashboards/my_dashboard/my_dashboard.yaml.

    This file contains the dashboard definition. It must conform to the JSON schema defined in ee/app/validators/json_schemas/analytics_dashboard.json.

  3. Optional. To create new visualizations to add to your dashboard, see defining a chart visualization.

For example, if you want to create three dashboards (Conversion funnels, Demographic breakdown, and North star metrics) and one visualization (line chart) that applies to all dashboards, the file structure looks like this:

.gitlab/analytics/dashboards
├── conversion_funnels
│  └── conversion_funnels.yaml
├── demographic_breakdown
│  └── demographic_breakdown.yaml
├── north_star_metrics
|  └── north_star_metrics.yaml
├── visualizations
│  └── example_line_chart.yaml

Define a chart visualization

You can define different charts and add visualization options to some of them, such as:

  • Line chart, with the options listed in the ECharts documentation.
  • Column chart, with the options listed in the ECharts documentation.
  • Data table.
  • Single stat, with the only option to set decimalPlaces (number, default value is 0).

To define a chart visualization for your dashboards:

  1. In the .gitlab/analytics/dashboards/visualizations/ directory, create a .yaml file. The filename should be descriptive of the visualization it defines.
  2. In the .yaml file, define the visualization configuration, according to the schema in ee/app/validators/json_schemas/analytics_visualization.json.

For example, to create a line chart that illustrates event count over time, in the visualizations folder create a line_chart.yaml file with the following required fields:

  • version
  • type
  • data
  • options

To contribute, see adding a new visualization render type.

Create a custom dashboard

To create a custom dashboard:

  1. On the left sidebar, select Search or go to and find your project.
  2. Select Analyze > Analytics dashboards.
  3. Select New dashboard.
  4. In the New dashboard input, enter the name of the dashboard.
  5. From the Add visualizations list on the right, select the visualizations to add to the dashboard.
  6. Optional. Drag or resize the selected panel how you prefer.
  7. Select Save.

Edit a custom dashboard

You can edit your custom dashboard’s title and add or resize visualizations in the dashboard designer.

To edit an existing custom dashboard:

  1. On the left sidebar, select Search or go to and find your project.
  2. Select Analyze > Analytics dashboards.
  3. From the list of available dashboards, select a custom dashboard (one without the By GitLab label) you want to edit.
  4. Select Edit.
  5. Optional. Change the title of the dashboard.
  6. Optional. From the Add visualizations list on the right, select other visualizations to add to the dashboard.
  7. Optional. In the dashboard, select a panel and drag or resize it how you prefer.
  8. Select Save.

Create a custom visualization

To create a custom visualization:

  1. On the left sidebar, select Search or go to and find your project.
  2. Select Analyze > Analytics dashboards.
  3. Select Visualization designer.
  4. In the Visualization title field, enter the name of your visualization.
  5. From the Visualization type dropdown list, select a visualization type.
  6. In the What metric do you want to visualize? section, select a measure or a dimension.
  7. Select Save.

After you save a visualization, you can add it to a new or existing custom dashboard in the same project.

Generate a custom visualization with GitLab Duo

Tier: For a limited time, Ultimate. On October 17, 2024, Ultimate with GitLab Duo Enterprise. Offering: GitLab.com Status: Experiment
History
  • Introduced in GitLab 16.11 as an experiment feature with a flag named generate_cube_query. Disabled by default.

Prerequisites:

To generate a custom visualization with GitLab Duo using a natural language query:

  1. On the left sidebar, select Search or go to and find your project.
  2. Select Analyze > Analytics dashboards.
  3. Select Visualization designer.
  4. In the Visualization title field, enter the name of your visualization.
  5. From the Visualization type dropdown list, select a visualization type.
  6. In the Generate with GitLab Duo section, enter your prompt. For example:

    • Daily sessions
    • Number of unique users, grouped weekly
    • Which are the most popular pages?
    • How many unique users does each browser have?
  7. Select Generate with GitLab Duo.
  8. Select Save.

After you save a visualization, you can add it to a new or existing custom dashboard in the same project.

Provide feedback on this experimental feature in issue 455363.

Visualization query builder

History
  • Introduced in GitLab 17.1 with a flag named analytics_visualization_designer_filtering. Disabled by default.
  • Generally available in GitLab 17.2. Feature flag analytics_visualization_designer_filtering removed.

You can use measures and dimensions to filter and refine the results of a custom visualization:

  • Measures: Properties that can be calculated. Measures are aggregated by default.
  • Dimensions: Attributes related to a measure. You can add multiple dimensions to a measure.

You can filter by custom event names with select measures:

  • Tracked events count
  • Tracked events unique user count
note
When you change or remove a measure then dependent dimensions may also be removed.

Troubleshooting

Something went wrong while loading the dashboard.

If the dashboard displays a global error message that data could not be loaded, first try reloading the page. If the error persists:

Invalid dashboard configuration

If the dashboard displays a global error message that the configuration is invalid, check that your configurations match the dashboard JSON schema defined in ee/app/validators/json_schemas/analytics_dashboard.json.

Invalid visualization configuration

If a dashboard panel displays a message that the visualization configuration is invalid, check that your visualization configurations match the visualization JSON schema defined in ee/app/validators/json_schemas/analytics_visualization.json.

Dashboard panel error

If a dashboard panel displays an error message:

  • Make sure your Cube query and visualization configurations are set up correctly.
  • For product analytics, also check that your visualization’s Cube query is valid.

Generate visualization with GitLab Duo returns unexpected results

If GitLab Duo doesn’t return the expected or a useful result, try editing your query to:

  • Specify a date range. For example: number of unique users in 2023 to 2024, grouped monthly.
  • Use the same names for metrics and dimensions as shown in the visualization designer. For example: returning users instead of existing customers.