Topic 07 of 07
Deduplication & matching
Find and merge records that refer to the same place or household despite messy spelling and formatting differences.
Why it matters
Duplicate addresses inflate costs, fragment customer history, and skew analytics. Matching turns noisy lists into a coherent view of places and people.
Exact vs. fuzzy
Exact matching on standardized keys is fast and safe but misses “123 Main St” vs “123 Main Street Apt 1” when unit handling differs.
Fuzzy and probabilistic matching weigh street tokens, phonetics, unit numbers, and occupant names. Tune thresholds to your tolerance for false merges.
Householding and entities
Decide whether you merge to a delivery point, a household, or a legal entity. A multi-unit building is one building and many deliverable addresses.
Keep survivorship rules explicit: which field wins, how history is retained, and how to unmerge when you get it wrong.
Prerequisites
Dedup quality tracks standardization and verification quality. Clean inputs dramatically reduce false positives and missed twins.
Related topics
- Standardization & formatting→
Rewrite addresses into a consistent, preferred format so systems can compare, store, and print them reliably.
- Address verification→
Confirm that an address exists in authoritative reference data and can typically receive mail or deliveries.
- NCOA — National Change of Address→
USPS-licensed process that updates addresses when people and businesses file official change-of-address moves.