Master data management is the practice of creating and maintaining one authoritative record for each core business entity, such as a property, tenant, or lease, and keeping that record consistent across every system that uses it. It reconciles conflicting copies into a single governed record called the golden record.
How Master Data Management Works
Master data management works by matching duplicate records that describe the same real-world entity, then merging their fields into one golden record under explicit rules. The process runs in four stages: identify the master entities, match records that point to the same one, apply survivorship rules to pick the winning value for each field, and govern the result so it stays trusted over time.
Matching is the hard step. Two systems rarely store the same property with the same string, so the engine compares addresses, parcel numbers, names, and other keys, scoring how likely two records are the same entity. High-confidence matches merge automatically; borderline pairs route to a human steward. According to DAMA International's Data Management Body of Knowledge (DAMA-DMBOK), the discipline that governs these golden versions and their replicas is Reference and Master Data Management.
Survivorship decides which value survives when sources disagree. A survivorship rule might prefer the most recently updated source, the most complete field, or a designated system of record for that attribute.
Stage | What happens |
|---|---|
Identify | Define which entities are master data (property, tenant, lease) |
Match | Score record pairs to find duplicates of the same entity |
Merge and survive | Apply survivorship rules to select the winning value per field |
Govern | Stewards review conflicts and maintain the golden record over time |
Why Master Data Management Matters
Master data management matters because a commercial real estate operator runs the same property, tenant, and lease across accounting, leasing, and asset management systems, and each holds a slightly different version. Without one authoritative record, square footage, ownership, and rent figures drift apart, and no report can be trusted end to end.
The cost of that drift is measurable. Gartner estimates that poor data quality costs organizations an average of $12.9 million per year in wasted work and lost opportunity. In real estate, the failure is concrete: a portfolio roll-up double-counts a building recorded under two spellings. Master data management is the discipline that makes "the property" mean one thing across the stack.
Example
Master data management is easiest to see when the same property sits in three systems under three names. An operator holds one asset in Austin, but each system stored it differently, so a naive count shows three properties and inflates the portfolio.
System | Property name | Square feet | Owner entity |
|---|---|---|---|
Accounting | 100 Congress Ave | 411,000 | Congress LP |
Leasing CRM | 100 Congress Avenue, Ste 2000 | (blank) | Congress Holdings LP |
Asset management | One Hundred Congress | 411,500 | Congress LP |
The matching engine compares address and parcel keys, scores the three records as the same entity, and links them. Survivorship rules then build one golden record: keep the standardized address, take square footage from the asset management system as the designated source (411,500), fill the blank field from whichever system has it, and resolve the owner entity to the value two of three sources agree on (Congress LP). The result is a single record with a stable ID that all three systems now reference. The portfolio count drops from three to one.
Variations and Edge Cases
Master data management behaves differently depending on how the golden record is stored and how strict the matching is. The table below covers the common architectural choices and where they break.
Variant | Behavior |
|---|---|
Registry style | Golden record is a virtual index of IDs; source systems keep their own copies |
Consolidation style | Golden records are compiled into a central hub for reporting |
Centralized style | The hub is authored first and pushed out to source systems |
Over-matching | Loose thresholds merge two different properties into one record |
Under-matching | Strict thresholds leave duplicates unmerged, defeating the purpose |
Master Data Management vs Data Governance
Master data management is often confused with data governance, but they operate at different levels. Master data management is the operational discipline that produces and maintains the golden record for each entity. Data governance is the policy layer that decides who owns the data, what the rules are, and how quality is enforced. One executes; the other authorizes.
NIST defines data governance as "a set of processes that ensures that data assets are formally managed throughout the enterprise," where the governance model "establishes authority and management and decision making parameters related to the data." Master data management is where those parameters get applied to a specific property or lease. Governance writes the survivorship policy; master data management runs it.
Master data management | Data governance | |
|---|---|---|
Scope | Master entities: property, tenant, lease | All data assets and policy |
Output | Golden records | Rules, ownership, accountability |
Question | What is the one true record? | Who decides, and by what rule? |
Frequently Asked Questions
What is a golden record in master data management? A golden record is the single authoritative version of an entity, built by matching duplicate records and applying survivorship rules to select the winning value for each field. It is the trusted record every system should reference instead of holding its own conflicting copy.
Is master data management the same as a single source of truth? No. A single source of truth is the goal, one trusted place for a given fact. Master data management is the practice that produces it through matching, survivorship, and governance. The single source of truth is the outcome; master data management is the method.
What counts as master data in commercial real estate? Master data is the core, slow-changing entities the business is built on: properties, tenants, leases, owners, and legal entities. Transactional data like rent payments and maintenance tickets is not master data, though it references the master records by their stable IDs.
Related Terms
Single Source of Truth