Reference data management
Reference Data Management
Give shared reference data an owner and a controlled change process.
Country codes, cost centres, currencies, classifications and organisational structures often start in a spreadsheet. Once several teams and applications depend on them, every change needs a clear owner, an agreed definition and a publication that consumers can identify.
Refrncr provides a single source for reference data, with controlled changes, published versions, review workflows and point-in-time access through an API.
The domain
What is reference data?
Reference data describes and classifies the information used across business processes, applications and reports. Common examples include country codes, currencies, cost centres, departments, GL accounts, product categories, tax classifications and error codes.
It changes relatively infrequently, but many systems depend on it being correct and consistent.
Unlike transactional data, reference data provides the context needed to interpret business information.
Departments
SALESFINOPSCost centres
410042004300Product categories
HWSWSVCShipping rates
NL 4.95BE 6.95DE 7.95Error codes
ERR-001WARN-204SYNC-503Countries
NLBEDEWhere it breaks
Why spreadsheets stop working
A spreadsheet is often the right place to start. The problems begin when multiple teams, systems and reports depend on the same reference data.
Without a system of record, the same dataset gradually diverges across applications, reports and spreadsheets.
Many copies
shipping_rates_v3_FINAL.xlsxrates_2026.csvSHIPPING_LOOKUPdim_shipping_rateRates (Sheet2)- Multiple copies
- No clear owner
- No review process
- No version history
- Different answers across systems
- Difficult historical reporting
A system of record
One source for shared reference data
Refrncr provides a dedicated system of record for reference data.
Each dataset has a defined schema, a named owner, validation rules, a review workflow and published versions that remain unchanged after release.
Changes are managed once, reviewed, published and made available consistently to every consuming system.
| Code | Destination | Rate |
|---|---|---|
| NL | Netherlands | 4.95 |
| BE | Belgium | 6.95 |
| DE | Germany | 7.95 |
Ownership
Treat reference data as a managed product
Reference data is often treated as a by-product of operational systems.
A more reliable approach is to treat each dataset as a managed product. That means a stable business key, a documented definition, a responsible owner and a controlled publication process.
Published versions remain unchanged while editors prepare the next release in a draft.
Controlled change
How the change process works
Reference data should not be edited directly in production. Changes move through a predictable process.
Set responsibility and definition
Choose the owner, define the schema and agree on the purpose of the dataset.
Prepare changes in a draft
Edit records or import CSV and Excel files. Validate and review the proposed changes.
Publish with a business date
Create an immutable published version with a release note and effective date.
Let consumers select a version or date
Consumers can request the latest version, a specific version or the version valid on a particular date.
Versions and audit
Know what changed and when
Every published version remains available.
Refrncr records who changed what, when it changed, who approved it and when it became effective.
Immutable versions, review decisions and audit history remain together.
Point-in-time reporting
Report using the data that was valid at the time
Reference data changes over time. Historical reports should use the data and structures that were valid during the reporting period, not whatever happens to be current today.
A cost centre moves from Sales to Operations on 1 June. A report generated on 15 July should still classify January to May under Sales, and June onwards under Operations.
Refrncr preserves both states so reports remain historically accurate and reproducible.
Three structures
Reference data is more than a code list
Organisations need more than a simple list of values. Reference data exists as datasets, mappings and hierarchies.
Datasets
Managed lists of reference values.
| Code | Destination | Rate |
|---|---|---|
| NL | Netherlands | 4.95 |
| BE | Belgium | 6.95 |
| DE | Germany | 7.95 |
Mappings
Relationships between systems.
Hierarchies
Structures used for reporting and aggregation.
- Organisation
- Commercial
- Sales
- Marketing
- Finance
- Operations
- Commercial
Scope
Do you need a full MDM platform?
Master Data Management and Reference Data Management solve different problems.
MDM platforms focus on mastering customers, products and other complex entities.
Refrncr focuses on governed reference data, mappings, hierarchies, versioning and historical reporting.
If your challenge is governed reference data, you may not need a full MDM implementation.
| Capability | MDM | refrncr |
|---|---|---|
| Golden records | ✓ | — |
| Matching | ✓ | — |
| Deduplication | ✓ | — |
| Reference data | ✓ | ✓ |
| Mappings | ✓ | ✓ |
| Hierarchies | ✓ | ✓ |
| Versioning | Varies | ✓ |
| Effective dating / point-in-time access | Varies | ✓ |
Both MDM platforms and refrncr can manage reference data. Refrncr focuses specifically on governance, versioning and effective dating for shared reference data.
Where to start
Common reference data use cases
Cost centre management
Manage cost centre ownership, reporting structures and organisational changes.
Country code management
Keep country codes consistent across ERP, CRM, reporting and integrations.
Currency management
Manage currency lists with version history and effective dates.
Product categorisation
Maintain consistent product classifications across applications and reporting.
System mapping
Manage SAP, ERP and CRM code translations without hard-coding mappings in applications.
Why refrncr
Why teams choose refrncr
- Single source for shared reference data
- Datasets, mappings and hierarchies
- Review before publication
- Immutable published versions
- Complete audit trail
- Point-in-time access
- Self-hosted and developer friendly
- Start with SQLite and scale to PostgreSQL
Start with real data
Install refrncr and try it with your own reference data
Start with a real dataset such as cost centres or country codes and publish it through the API in minutes.
Start with a dataset that already has multiple consumers so you can experience ownership, versioning and publication in a realistic scenario.