4.1 Creating and Editing Tables in the Data Workspace

Every intelligent application built on the Power Platform stands on a data model, and in Microsoft Dataverse that model is made of tables. For AB-410, you must know exactly where tables are created, which creation method fits a given scenario, and where the visual designer stops.

The Tables Area and the Data Workspace

Sign in to make.powerapps.com, pick your environment, and select Tables in the left navigation. Three tabs filter the list: Recommended (standard tables shipped with Dataverse or Dynamics 365 apps), Custom (tables you and other makers created), and All.

Selecting New table drops you into the data workspace — a visual canvas where tables appear as draggable cards, rows can be edited inline in the table row editor, and relationships form a live entity-relationship diagram. Command-bar actions to know by name:

• New table — create a table with any creation method
• Existing table — add tables already in the environment to visualize them
• Create relationships — select one table, CTRL + click a second, then define the relationship
• Row ownership — switch ownership between user/team and organization
• Fit view — re-center the diagram when tables seem lost

Exam trap — workspace limitations

The workspace is fast but not feature-complete. Column types not supported there are Rich text, Customer, Autonumber, and Formula. And while new-to-new one-to-many and many-to-one relationships work, many-to-many relationships cannot be created in the workspace, nor can most combinations involving existing tables. If a scenario asks why a maker cannot add an autonumber column in the designer, the answer is to open the table hub or solution explorer instead.

Four Ways to Create a Table

Copilot-assisted creation

With Start with Copilot, you describe the data you need — for example, 'Track building inspections, the inspector assigned, and the inspection result' — and Copilot generates one or more tables, suggested columns, sample rows, and even relationships. The Table options menu controls the number of tables, table size, and whether relationships are included; keep iterating in the Copilot panel, then select Save and exit. Copilot is enabled by default in environments and the experience is available in 20 languages.

Exam trap

Copilot drafts the schema; it does not make architectural decisions for you. You still own the ownership type, column data types, and required levels — which is exactly where the exam goes next.

Standard Tables: Extend Before You Build

Dataverse ships with hundreds of standard tables. The exam favorites:

• Account — companies or organizations you do business with
• Contact — individual people
• User — people licensed in the environment
• Activity — the base for task-like records such as email, phone call, and appointment

The golden rule: extend a standard table when it already models your concept. Add custom columns to Account rather than cloning it — you inherit forms, views, relationships, search, and the activity timeline for free. Create a custom table only when no standard table fits, such as Inspection or Certification. Custom schema names carry the solution publisher's customization prefix (for example, contoso_Inspection).

Ownership Types

Activity tables can only be user or team owned — never organization-owned. Ownership also drives which security-role privileges apply, so exam questions love swapping the two.

Table Types (Awareness Level)

• Standard — the default; data stored in Dataverse
• Activity — tasks and appointments with times and participating parties
• Virtual — rows live in an external source such as Azure SQL or SharePoint, surfaced as Dataverse tables without replication
• Elastic — for very large data volumes needing high throughput, storage, and low latency; think IoT telemetry

Microsoft's own guidance is to consider elastic tables when the scenario entails very large data volumes with high throughput and low-latency requirements — that phrasing shows up on exams.

The Primary Name Column

Every table gets a primary column automatically — display name usually Name, a single line of text. Lookups pointing at the table display this value, so make it human-readable ('Inspection 1042', not a GUID). You can change its display name and description during creation on the Primary column tab, but after the table is saved they are locked, and its data type can never change.

Importing from Excel and SharePoint

Create with external data uploads an Excel/.CSV file or connects to a SharePoint list, and Copilot infers column names, data types, and a primary column. Only the first 20 rows preview in the designer; the full dataset ingests after you save. Some SharePoint column types do not map — including Image, Task outcome, Managed metadata, External data, and Attachments — so plan a cleanup pass.

Finally, remember the destructive edge: deleting a custom table deletes all of its data and any child rows in parental relationships, and the only recovery is restoring an environment backup.