Choosing a master data management architecture is one of those decisions that looks technical but is really about fit. Pick the wrong style and you either over-engineer a simple problem or under-build a critical one. Pick the right one and MDM quietly does its job for years.
Reference and master data is one of the knowledge areas defined in the DAMA DMBOK, and the architecture you choose shapes how well the rest of your data management works. There are four main MDM implementation styles: registry, consolidation, coexistence and centralised. They are not ranked from worst to best. Each suits a different mix of sources, use cases and appetite for change. This article explains what each one does, where it fits, and how to choose.
The four styles at a glance
- Registry: links records across systems without moving the data. Lightest touch.
- Consolidation: pulls data into a hub to build a golden record for reporting and analytics.
- Coexistence: master data lives in both the sources and the hub, kept in sync.
- Centralised: the hub becomes the single authoring point for master data. Strongest control.
What MDM architecture actually decides
Master data is the shared data every system relies on: customers, products, suppliers, accounts. MDM architecture decides three things: where the trusted version lives, whether that version can be written back to source systems, and how much your existing systems have to change. Everything else follows from those three answers.
Style 1: Registry
A registry style leaves master data where it is. It stores only the cross-reference keys and a thin matching index that links the same customer or product across different systems. The golden record is assembled on demand by reading from the sources, rather than stored as a single physical record.
Best for: a unified view quickly, across many systems, with minimal disruption. Common when you cannot or do not want to change the source systems.
Trade-off: it identifies and links, but it does not clean or standardise the underlying data, and there is no single place that owns a corrected record.
Style 2: Consolidation
A consolidation style pulls master data from the sources into a central hub, matches and merges it, and builds a golden record there. The sources stay authoritative for day-to-day work. The golden record is used downstream, mainly for reporting, analytics and a trusted 360-degree view.
Best for: business intelligence, analytics and reporting where you need one clean version, but you do not need to push corrections back into operations.
Trade-off: it is usually batch, so the golden record is only as current as the last load, and it does not fix data at the source.
Style 3: Coexistence
A coexistence style maintains the golden record in the hub and keeps it synchronised with the source systems in both directions. Data can be authored in the sources or in the hub, harmonised centrally, then shared back. It sits between consolidation and full centralisation.
Best for: organisations that want a trusted, maintained master record and the ability to feed it back into operational systems, without forcing every application to author data in one place.
Trade-off: two-way synchronisation is more complex to build and run, and it needs mature governance to avoid conflicts.
Style 4: Centralised
A centralised style, sometimes called a transaction style, makes the hub the single system where master data is created and maintained. Applications consume master data from the hub rather than holding their own versions. This gives the strongest control and the cleanest, most consistent master data.
Best for: high-governance environments, regulated industries, and cases where consistency of master data is critical to operations or compliance.
Trade-off: it is the most invasive to implement, because existing systems have to change how they source master data, and it demands the most governance and change management.
How to choose the right MDM style
A quick comparison of the four against the questions that matter most:
| Style | Golden record | Writes to source | Typical latency | Effort | Best for |
|---|---|---|---|---|---|
| Registry | Assembled on demand | No | Real-time read | Low | Fast unified view, many sources, no source changes |
| Consolidation | In the hub | No | Batch | Medium | Reporting, analytics, 360 view |
| Coexistence | In the hub, synced both ways | Yes | Near real-time or batch | High | Trusted record fed back to operations |
| Centralised | In the hub only | Sources consume from hub | Real-time | Highest | Strong governance, regulated data |
Then weigh the practical questions: how many source systems you have, whether corrections need to flow back into operations, whether analytics or operations is the priority, how much change your systems can absorb, and how mature your governance is. Governance maturity matters more than people expect. The more control a style gives, the more governance discipline it needs to work.
The most common mistake
The most common mistake is choosing a style by ambition rather than need. Teams reach for a centralised hub because it sounds like the most complete answer, then stall under the weight of the change it demands. Many organisations are better served by starting with a registry or consolidation style, proving value on one or two domains such as customer or product, and moving towards coexistence or centralisation only when the governance and the business case are ready.
Tooling follows the same logic. Whether you build on Informatica MDM or another platform, the platform should serve the style you have chosen, not decide it for you.
Frequently asked questions
What are the four MDM implementation styles?
Registry, consolidation, coexistence and centralised. They differ in where the golden record lives, whether it writes back to source systems, and how much existing systems must change.
Which MDM style is best for analytics?
Consolidation is usually the best fit. It builds a clean golden record in a hub for reporting and a trusted 360-degree view, without changing operational systems.
Which MDM style gives the most control?
The centralised or transaction style, where the hub is the single place master data is created and maintained. It offers the strongest consistency but is the most demanding to implement.
Do we have to pick one style forever?
No. Many organisations start light with a registry or consolidation style and evolve towards coexistence or centralisation as governance matures and the business case grows.
Choosing an MDM approach?
We offer a free consultation to help you match the right MDM style to your sources, your use cases and your governance maturity, before you commit to a platform.
Reach us at info@dheerayatsolutions.com or on WhatsApp at +91 83369 23288.
CDMP, DAMA, and DMBOK are trademarks of DAMA International. This course is an independent training programme aligned to DMBOK v2 and is not official DAMA material.