Address Cleansing

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