A connected table pulls record-level data, such as your employee roster, straight from an HRIS your Cube team has enabled (Paylocity, Rippling, Justworks, and others) into a Cube table. You pick the fields once, set a filter, map departments and locations to your Cube dimensions, and refresh on demand with Import. Every row you add and every column you create stays put through each refresh.
What you can use it for
- Headcount planning without the export. Bring the live roster into a table, add your own columns for merit, bonus, or planned changes, and let formulas read them with TABLE_SUM and TABLE_LOOKUP.
- Keeping scenarios in step. Import into Forecast and Budget at the same time, or freeze a scenario once the plan is approved.
- Filtering to what matters. Import only active employees, or only one entity, and keep dimension values you do not plan for out of the table.
- One source, many tables. Create a roster table and a contractor table from the same integration, each with its own fields, filters, and mapping.
Before you start
- You need Integrations access and permission to create tables. Admin user types have both by default.
- Your Cube Implementation Manager or Support must enable the integration for Planning Tables and turn on the dimensions you want to map (for example Location and Cost Center). If the integration does not appear on the Data Source step, that is the first thing to check.
- Decide which scenarios the table should live in. Scenarios set to Write Protection: Blocked cannot be selected.
- Each connected table owns its own columns, filter rules, dimension mapping, and import history. The integration login is the only thing tables share, so settings on one table never affect another.
How to create the table
1. In the Workspace, go to Tables and select + Table.
2. On Data Source, select Connect next to the integration to pull from. To build a table by hand instead, select Start from scratch.
3. On Select Source Fields, check the fields to bring into the table, then select Next. Fields tagged Enabled dimension are the ones you can map to a Cube dimension.
4. On Configure Columns, rename any column and set its Column Type, then continue to the next step. Types are pre-filled from the source field.
5. On Apply Data Filters, select + Add rule for each field you want to filter on: select the field, select is in or is not in, type a value, and press Enter. Select Next.
Screenshot 6: Data Filters step with two rules, Employment Status is in FULL_TIME and Pay Frequency is in YEARLY
6.On Review, enter the Name, select the scenarios under Save to Scenarios, set Access, check the filters and columns, and select Create.
The table opens empty with the message "This table is connected to a source but hasn't been refreshed yet. Click Import to bring in your records." Cube syncs the dimension values from the source in the background.
If the table has a dimension column, map its values first (see Map Dimension Columns in a Connected Table), then select Import. A table with no dimension column can be imported straight away.
Choosing column types
| Column Type | Use it for | What happens on Import |
|---|---|---|
| Text | Names, titles, codes, anything you will not calculate with | The source value is written as text |
| Numeric | Salary, hours, rates | The source value is written as a number. Values that cannot be read as a number fail the Import. |
| Date | Start date, termination date | The source value is written as a date |
| Dimension | Fields tagged Enabled dimension, such as Location or Cost Center, that you want to roll up by a Cube dimension | Each source value is written as the Cube member you mapped it to. Values mapped to Ignore are not brought in. |
Only fields that match a dimension enabled on the integration can be dimension columns, and they default to that type. Fields such as employee name or Employee Number can only be Text, Numeric, or Date. You can switch a dimension field to Text if you do not want to map it.
How filter rules work
Filter rules decide which source records the table admits. Each rule names a field, an operator (is in or is not in), and one or more values. Values are free text and match regardless of case or surrounding spaces; dates are compared as dates. All rules must match for a record to come in ("Include records when all match"), and each field can carry one rule. With no rules, every record is imported.
Rules apply to Text, Numeric, and Date fields only. Dimension fields do not appear in the field list: to keep out a location or cost center, map it to Ignore on the table's Mapping page instead. The two behave the same way, so a record that never qualified is never brought in.
Examples
Active, salaried employees only
A finance manager creates a Paylocity Roster table, keeps Display Full Name, Employee Number, Employment Status, Start Date, Pay Rate, Location, and Cost Center 1, and adds two rules: Employment Status is in FULL_TIME and Pay Frequency is in YEARLY. Location and Cost Center 1 stay as dimension columns. After mapping six locations and seven cost centers to Cube members, she selects Import. 305 employees arrive as Synced rows; hourly and part-time staff never appear.
Leaving out a cost center
The same manager does not plan for contractors, who sit in a CONTR cost center at the source. Instead of a filter rule, she maps CONTR to Ignore on the Mapping page. The contractor records are never brought in, and the table rolls up cleanly by Cost Center.
Two tables, one integration
An FP&A lead creates a second table from the same Rippling integration for a contractor forecast, selecting different fields and a different filter. The two tables share only the login. Mapping Location on the roster table does nothing to the contractor table, so she maps it separately.
If something is not right
| What you see | What to check |
|---|---|
| Your HRIS is not listed on the Data Source step | The integration is not enabled for Planning Tables, or you do not have Integrations access. Contact your Implementation Manager or Support. |
| Your first Import fails | Open the Mapping page and check for unmapped values. Then check that your filter values match the source exactly (for example YEAR rather than YEARLY). |
| A scenario is greyed out on Review | It is set to Write Protection: Blocked. Connected tables cannot be added to Blocked scenarios. |