Most organisations hold the same customers, products and suppliers in several systems — ERP, CRM, e-commerce and finance. When those records disagree, reports conflict, processes break and customers notice. Master data management (MDM) addresses the problem directly.
What counts as master data?
Master data describes the core entities a business runs on: customers, products, suppliers, employees, locations and accounts. It changes relatively slowly and is shared across many processes, unlike transactional data such as orders and invoices.
Common MDM styles
- Registry: links records across systems without changing them; quick to start.
- Consolidation: builds a cleaned “golden record” for reporting.
- Coexistence: keeps a golden record and synchronises improvements back to sources.
- Centralised: creates and maintains master data in one hub that feeds all systems.
The core capabilities
MDM relies on matching and merging duplicate records, data quality rules, a governance workflow for changes, and integration with source systems. Technology helps, but agreed definitions and data ownership matter just as much.
Getting started
- Choose one domain with a clear business pain — often customer or product.
- Agree definitions and the attributes that make up the golden record.
- Profile source data to understand duplicates and gaps.
- Start with a registry or consolidation style, then mature.
5 best practices for successful MDM
- Secure business sponsorship. MDM changes processes across sales, finance and operations, so it needs an executive sponsor who owns the outcome.
- Define the golden record precisely. Agree which attributes matter, which source is trusted for each attribute and how conflicts are resolved.
- Fix data at the source. Cleansing data downstream helps reporting, but lasting improvement comes from better validation and workflows where records are created.
- Automate matching with human review. Use matching rules and machine learning to identify duplicates, with stewards reviewing uncertain matches.
- Measure quality over time. Track duplicate rates, completeness and time to create new records, and report improvements to stakeholders.
Business benefits
- Accurate customer views for sales, service and marketing.
- Faster product launches with consistent product information across channels.
- Reduced supplier risk and duplicate payments.
- Reliable reporting and easier regulatory compliance.
Common mistakes to avoid
- Treating MDM as a one-off data cleansing project.
- Trying to master every domain simultaneously.
- Ignoring the integration effort needed to keep systems synchronised.
- Leaving data stewards without the authority to resolve conflicts.
Frequently asked questions
How is MDM different from a data warehouse?
A data warehouse stores historical data for analysis. MDM creates and maintains the trusted version of core entities that operational systems and the warehouse both rely on.
How long does an MDM programme take?
A first domain often takes six to twelve months to deliver meaningful results, with further domains added incrementally.
A 90-day action plan
Days 1 to 30: choose the customer, product or supplier domain, quantify the business pain caused by duplicates and inconsistencies, and secure a sponsor.
Days 31 to 60: profile records from the main source systems, agree the attributes and survivorship rules for the golden record and appoint stewards.
Days 61 to 90: implement matching for one region or business unit, publish quality metrics and plan integration back to operational systems.
Questions to ask vendors
- Which hub styles do you support, and can we evolve between them?
- How configurable are matching and merging rules?
- What steward workflows and audit trails are included?
- Which connectors exist for our ERP, CRM and e-commerce systems?
- How do you handle hierarchies, such as parent companies and product families?
Key terms explained
- Golden record: the single trusted version of an entity created from several sources.
- Survivorship: rules that decide which source value wins when sources disagree.
- Match and merge: identifying duplicate records and combining them.
- Reference data: standard code lists, such as country or currency codes.
- Hierarchy management: maintaining parent and child relationships between records.
The bottom line
Trusted core records are the foundation for accurate reporting, efficient operations and good customer experiences. Success requires business sponsorship, precise definitions, fixing problems where records are created, automated matching with human review and ongoing quality measurement. Start with one domain that has a clear business problem, deliver visible improvements and expand step by step. Technology helps, but ownership and agreed definitions make the difference.
Further reading on master data management
For authoritative, vendor-neutral guidance on master data management, see DAMA International. You can also browse our free whitepapers.
