The foundation of unified data
Master Data Management sits at the intersection of data quality and data governance. At its core, MDM is about creating a single, authoritative view of critical business entities — customers, products, suppliers — by matching and merging records scattered across systems.
Deterministic matching
Exact and rule-based approaches for high-confidence record linking. Match on email addresses, phone numbers, national IDs, or composite keys with zero tolerance for ambiguity.
Probabilistic matching
When exact matches aren’t available, fuzzy matching algorithms bridge the gap:
- Soundex and Metaphone — phonetic similarity for names
- Jaro-Winkler — string similarity weighted toward prefix matches
- Levenshtein distance — edit-distance-based comparison
Merge strategies
Once matches are identified, survivorship rules determine which values win:
- Recency-based — the latest value wins
- Source authority — trusted systems take precedence
- Frequency-based — the most common value across sources
The connection to CDP identity resolution
The same matching and merging principles that power MDM directly apply to CDP identity resolution — the algorithms are identical, just applied to customer profiles instead of master records.
Read the full article on Medium.