Atlas logo
⌘K

Documentation

Getting Started
Data
Maps
Builder
Workflows
Forms
Sharing & Collaboration
Projects & Management
Connections

Relations & Lookups

Link datasets together and access values from related records

Relations and Lookups allow you to connect datasets together, creating powerful data models without duplicating information. Link records across datasets and automatically pull in related values—keeping your data organized, consistent, and always up-to-date.

Understanding Relations

A Relation field creates a link between records in two different datasets. This allows you to:

  • Connect related data without duplicating it
  • Build hierarchical data structures (e.g., projects → tasks, customers → orders)
  • Create many-to-one or many-to-many relationships
  • Keep information synchronized across datasets

Example Use Cases

ScenarioPrimary DatasetRelated DatasetRelationship
Project managementTasksProjectsMany tasks → One project
Customer ordersOrdersCustomersMany orders → One customer
Asset trackingAssetsLocationsMany assets → One location
InventoryProductsCategoriesMany products → One category
Field surveysObservationsSitesMany observations → One site

Adding a Relation Field

  1. Open your dataset in the data table view
  2. Click + Add Field in the table header
  3. Select Relation from the field type dropdown
  4. Configure the relation:
    • Related dataset: Select the dataset you want to link to
    • Display field: Choose which field to show when selecting records
  5. Click Create to save

Once created, you can click on any cell in the Relation field to:

  • Search for and select existing records from the related dataset
  • Link multiple records (for many-to-many relationships)
  • View linked record details
  • Navigate directly to linked records

Understanding Lookups

A Lookup field retrieves and displays values from a field in a linked record. Once you've established a relation, lookups let you bring specific data from the related dataset into your current dataset—without copying or duplicating it.

Lookups are read-only and automatically update when the source data changes.

How Lookups Work

  1. A Relation field connects Dataset A to Dataset B
  2. A Lookup field in Dataset A displays a specific field value from the linked record in Dataset B
  3. When the value in Dataset B changes, the Lookup in Dataset A updates automatically

Example

Scenario: You have a Tasks dataset linked to a Projects dataset.

Tasks DatasetField TypeValue
Task NameText"Design homepage"
ProjectRelation"Website Redesign"
Project StatusLookup"In Progress"
Project ManagerLookup"Sarah Chen"

The Lookup fields automatically show the Status and Manager from the linked Projects record.

Adding a Lookup Field

  1. Open your dataset in the data table view
  2. Click + Add Field in the table header
  3. Select Lookup from the field type dropdown
  4. Configure the lookup:
    • Source relation: Select an existing Relation field in your dataset
    • Field to lookup: Choose which field from the related dataset to display
  5. Click Create to save

Note: You must have at least one Relation field in your dataset before you can create a Lookup field.

Lookup Field Behavior

Multiple Linked Records

When a Relation field contains multiple linked records, the Lookup displays all corresponding values as a comma-separated list.

Relation FieldLookup (Status)
Project A, Project B"Active, Completed"

Empty Relations

If no record is linked in the Relation field, the Lookup field displays empty/blank.

Supported Source Field Types

Lookups can retrieve values from most field types in the related dataset:

  • Text, Number, Currency, Percentage
  • Date, Time, Duration
  • Single select, Multi select
  • Checkbox, Rating
  • Email, Phone, URL
  • User fields

Lookups cannot be created from other Lookup fields or Geometry fields.

Working with Relations in Formulas

You cannot directly reference Relation or Lookup fields in Formula fields. To perform calculations on related data, use the Lookup field values as inputs.

Example: Calculate a value based on looked-up data

IF(@Project Status == "Active", @Priority * 2, @Priority)

Where @Project Status is a Lookup field.

Data Integrity

Deleting Linked Records

When you delete a record that is linked from another dataset:

  • The Relation field in the source dataset becomes empty
  • The corresponding Lookup fields become empty
  • No data is lost in the source dataset itself

Changing Relations

When you change which record is linked:

  • All Lookup fields automatically update to reflect the new linked record's values
  • Previous values are not stored—Lookups always show current data

Best Practices

Designing Relations

  • Plan your data model before creating relations. Identify which datasets need to connect.
  • Use descriptive display fields so users can easily identify records when linking.
  • Keep relations simple when possible. Complex many-to-many relationships can be harder to maintain.

Using Lookups Effectively

  • Avoid duplicating data that could be looked up. If data exists in a related dataset, use a Lookup instead of copying it.
  • Create Lookups for frequently accessed fields to save time navigating between datasets.
  • Remember Lookups are read-only. To edit the value, navigate to the source record.

Common Patterns

Parent-Child Relationship

Connect child records to a single parent:

Projects (parent)

  • Project Name
  • Status
  • Manager

Tasks (child)

  • Task Name
  • Project → Relation to Projects
  • Project Status → Lookup from Projects

Reference Data

Link to a dataset of standardized values:

Inspections

  • Location
  • Asset Type → Relation to Asset Types
  • Inspection Interval → Lookup from Asset Types
  • Checklist Template → Lookup from Asset Types

Asset Types (reference)

  • Type Name
  • Inspection Interval
  • Checklist Template

Geographic Hierarchy

Connect locations to administrative boundaries:

Sites

  • Site Name
  • Site Location
  • Region → Relation to Regions
  • Regional Manager → Lookup from Regions

Troubleshooting

IssueSolution
Can't create LookupEnsure you have at least one Relation field in your dataset first.
Lookup shows blankCheck that the Relation field has a linked record. Verify the source field has a value.
Lookup shows multiple valuesThis is expected when the Relation links to multiple records. Each value is comma-separated.
Can't edit Lookup valueLookups are read-only. Navigate to the linked record to edit the source value.
Deleted record still showingRefresh your browser. The Relation should clear when the linked record is deleted.
Relation options not loadingCheck your permissions. You need access to the related dataset to link records.

Relations vs. Duplicating Data

ApproachProsCons
Relations + LookupsData stays in sync automatically; no duplication; single source of truthRequires more upfront planning; lookup values are read-only
Copying dataSimple to set up; values are editableData can become out of sync; storage waste; maintenance burden

Recommendation: Use Relations and Lookups whenever data logically belongs to another dataset and should stay synchronized.

PreviousTroubleshooting GuideData
NextDrawing ToolsMaps