Master Data Management (MDM): The Single Source of Truth
Master Data Management (MDM) creates a single source of truth for core business entities. It's used when departments have conflicting data (e.g., 'ACME Inc.' vs 'Acme Corp').
THE MENTAL MODEL: Master Data Management (MDM) is a discipline, not just a tool, for creating a single source of truth for an organization's most critical data assets. Think of it as creating one official, shared contact list for core business 'nouns' like Customer, Product, or Supplier. Without MDM, the sales CRM might list 'ACME Corp.', while the finance ERP has 'Acme Corporation, Inc.', leading to reporting errors and operational chaos. MDM's goal is to create one 'golden record' that all systems trust and use.
HOW IT WORKS: MDM is a collaborative process between business and IT. First, the organization identifies its master data domains (e.g., Customer, not transactional data like Invoice). Second, data is collected from all source systems. Third, cleansing and matching rules are applied to de-duplicate and consolidate records into a single, high-quality golden record. Fourth, data stewards—people assigned accountability for specific data domains—are appointed to govern the data's quality. Finally, this master record is syndicated, either by pushing it back to source systems or providing it via a central API to ensure uniformity across the enterprise.
WHEN TO USE IT: MDM is essential for large, complex enterprises with many disconnected applications. Use it when inconsistent data causes business problems, such as inaccurate financial reporting, failed product shipments due to wrong addresses, or poor customer experience. It's a prerequisite for reliable, enterprise-wide analytics.
WHEN NOT TO USE IT: MDM is overkill for small organizations that run on a single, monolithic application. It is not for managing transactional data (like web logs or orders) or unstructured data. Most importantly, do not attempt an MDM initiative if there is no executive sponsorship or willingness to enforce data governance policies; it will fail due to political resistance.
ONE CANONICAL EXAMPLE: A global bank wants to understand a customer's total risk exposure. The customer has a checking account in one system, a mortgage in another, and an investment portfolio in a third, each with slight variations in name or address. An MDM system consolidates these three profiles into one golden 'Customer' record. This allows the bank to see the customer's complete relationship and accurately calculate total risk, something impossible when the data was siloed and inconsistent.
Read the original → en.wikipedia.org
Get five bites like this every day.
Tezvyn delivers a daily feed of 60-second tech bites with quizzes to lock in what you learn.