Master data, and what Grunddata is not
If you work with master data, Grunddata can look like an MDM tool at first glance. It holds business data, it has owners and history, and other systems read from it. But it is not an MDM tool, and it does not try to be one. This guide explains what master data and MDM are, and which data is left over once they have done their job.
What master data is
Section titled “What master data is”Gartner defines master data as “the least number of consistent and uniform sets of identifiers and extended attributes that uniquely describe the core entities of an enterprise and are used across multiple business processes.”
In a bank, those core entities are usually parties (customers and counterparties), products, the internal organisation, and geography. Contracts, loans and transactions are not master data. They are transactional data that refer to it.
We use a bank as the example in this guide. Other industries have the same layers, with other names.
What MDM does
Section titled “What MDM does”Master data management is the discipline of keeping that master data consistent. In practice, an MDM tool does three things:
- Matching. It finds records in different systems that describe the same entity.
- Deduplication. It merges or links those records so that one customer is not counted as three.
- Golden record. It produces one agreed version of each entity, using rules for which source wins.
This is hard work, and it matters. It is also a specific job, done for a small number of entity types.
Why many banks have no single MDM
Section titled “Why many banks have no single MDM”The picture of one central MDM feeding every system is less common than it sounds. In many of the banks we talk to, each major system is the master for its own domain. The core banking system for accounts, the CRM for customer contacts, the HR system for the organisation. Integration and the data warehouse connect them.
That works for the core entities. It leaves a gap for everything else.
The data that is left over
Section titled “The data that is left over”Next to master data and transactional data sits a layer of data that belongs to neither:
- Classifications the business adds to parties and products, such as an internal risk group or segment.
- Reporting structures and hierarchies.
- Mappings, for example from product code to report line.
- Limits and thresholds.
- Parameters that feed models and calculations.
- Control data and the business’s own key figures.
No system created this data. The business did. It is often critical to reporting, but it has no master. It lives in spreadsheets, in tables IT updates on request, or in code.
Supplementary master data
Section titled “Supplementary master data”We call this layer supplementary master data. It supplements the master data in your core systems and MDM with what the business knows and decides. It is owned by the business, changes when the business changes, and needs the same care as any other data a report depends on.
Reference data, such as currency and industry codes, often sits alongside it. We cover that in our guide Reference data management.
What Grunddata is, and what it is not
Section titled “What Grunddata is, and what it is not”Grunddata is a business support system for reference data, supplementary master data and other business data. It is the master for data that has no other master. The business maintains the data in Excel, Google Sheets or the web app. Every change is validated and recorded with author, timestamp and message. Other systems read it through API and OData.
Grunddata is not an MDM tool. It does not match records across systems, does not deduplicate, and does not build golden records. It does not replace your MDM, your core systems or your data warehouse. It sits next to them.
How they work together
Section titled “How they work together”A simple example. Your MDM or core system holds the counterparty and its identifier. The risk team decides which internal risk group each counterparty belongs to. That classification lives in Grunddata, keyed on the same identifier. The data warehouse reads both and joins them. Each system does the job it was built for, and the business owns the part that is theirs.
Where to start
Section titled “Where to start”Look at one important report and list the data it depends on. Mark what comes from core systems, what comes from MDM, and what someone maintains by hand. The last group is supplementary master data. That is where to start.
Get in touch if you want to map this layer in your own reporting.