Skip to main content
Dimensions are the axes your data rolls up by: department, entity, account, product, channel, region, and revenue type. This is where admins define the canonical set and map raw source values onto it, so every filter, group-by, and pivot across the platform agrees.

​
Canonical dimensions

A canonical dimension is the single agreed-upon list Novaplan plans and reports against, independent of what any one source calls things. Each source then maps its raw values onto that list, so a department named one way in your HRIS and another way in your ledger still rolls up to the same canonical department.

​
Raw to canonical mapping

1

Define the canonical list

Maintain the canonical values for each dimension. This is the list everything rolls up to.
2

Map raw source values

For each source, map its raw values (a CRM stage, a source department name, a ledger entity) onto the canonical value. Mappings are stored in a registry, not hard-coded, so adding a source value is a configuration change, not an engineering one.
3

Reconcile entity names to IDs

External names (CRM accounts, HiBob employees, ledger vendors) reconcile to canonical IDs through a name-to-ID lookup that tries an exact ID match first, then a mapped name, then a confident fuzzy match. This is what keeps identity stable across syncs.
Identity resolves by ID, not by name. When a source renames a department or vendor, the mapping keeps your hierarchy stable instead of scrambling it. Each data page links back here through a “Manage mappings” entry point.

​
Renaming a value everywhere

In the Dimension Mapper grids, any cell that holds a label shared across multiple rows shows a small chip with a count (how many rows use that value). Click the chip - or right-click any cell carrying that value - and choose Rename everywhere to change it in every row at once. A confirmation dialog names the current value, the new value, and the count of rows that will be updated. This keeps your hierarchy consistent without having to hunt down every occurrence. To change just one row, edit the cell directly. Your new text becomes private to that row; the shared value stays unchanged on all other rows.

​
Solve

The Solve button analyzes your existing vocabulary and pre-fills rows whose target value already appears elsewhere in the mapping. When all gaps resolve this way, you can review the pre-filled rows and confirm without any manual entry. When some remain unfilled, Solve tells you exactly how many it resolved and how many still need attention.

​
Site Aliases

The Site Aliases tab (under Dimensions) maps raw location strings from your HRIS onto the site names your fringe, merit, commission, and headcount-fee configuration uses. Without this mapping, employees whose HRIS location does not exactly match a configured site name fall back to the global catch-all rate - which may be materially wrong for employees in countries with a per-country schedule already configured. A detector panel above the grid lists every HRIS site value that currently resolves to no configured fringe rate, with the employee count and the estimated cost at the fallback rate. Start there: the panel shows the impact of each unmapped site so you can prioritize which aliases matter most. Mapping takes effect at read time - no re-ingest required. Remove or change an alias and the next forecast picks up the correction automatically.
Site Aliases use exact matching only, not fuzzy matching. If your HRIS sends both “US” and “United States” and your config has different wage bases for each, Novaplan maps each string independently rather than collapsing them. Check the detector panel after adding aliases to confirm all high-value sites resolve.

​
Common questions

It is unmapped. Add the raw-to-canonical mapping for that source, then re-run curation. Until it is mapped, the value will not roll up to a canonical dimension.
Map both raw values onto the same canonical value. They then reconcile to one ID and roll up together everywhere.
Check the Site Aliases tab. If the HRIS location string does not exactly match the site name in your fringe config, those employees fall back to the global rate. Add an alias from the HRIS value to the correct config name, and the next forecast will use the per-site schedule.

​
Troubleshooting

  • A filter or pivot shows a value twice: two raw values are reconciling to two IDs. Map both onto one canonical value.
  • An account is missing from a report: confirm its source-to-canonical account mapping in GL Mapping.
  • Numbers do not reconcile after a rename: the rename should be absorbed by the ID-based mapping. If it is not, the source value is likely mapped by name; correct the mapping.
  • Fringe cost changed unexpectedly after a site alias was added: the alias corrected the rate for employees who were previously falling through to the global catch-all. Check the Site Aliases detector panel for the before/after estimate.