What is record matching?
Record matching is the underlying capability that powers dedupe, lead-to-account matching, and merges - deciding when two records are really one entity.
It blends exact matching on strong keys with fuzzy comparison, because real-world data rarely matches perfectly.
Short answer
Record matching is determining whether two records represent the same real-world entity, using exact keys and fuzzy comparison of fields like name, email, and company. It is the engine behind deduplication, lead-to-account matching, and merging. Good record matching combines strong identifiers like email and domain with fuzzy logic to catch the messy duplicates exact matching misses.
Key takeaways
- Determines whether two records are the same entity.
- Combines exact keys with fuzzy field comparison.
- Powers dedupe, lead-to-account matching, and merges.
- Strong identifiers plus fuzzy logic catch real duplicates.
Why it matters
Matching is the foundation under several data-ops capabilities. Get it right and dedupe, lead-to-account matching, and merges all work; get it wrong and you either miss duplicates or wrongly merge distinct records. The quality of matching determines the quality of everything built on it.
How Ardovo handles it
Ardovo matches on strong identifiers like email and domain plus fuzzy comparison, scoring each candidate by confidence. Rook auto-acts on high-confidence matches and flags borderline ones, so dedupe, lead-to-account matching, and merges all rest on reliable matching.
Frequently asked questions
What is record matching used for?
Determining whether two records are the same entity, which powers deduplication, lead-to-account matching, and record merging. Any capability that needs to know "are these the same person or company" relies on record matching underneath, making it a foundational data-ops function.
How does record matching work?
It combines exact matching on strong identifiers - email for people, web domain for companies - with fuzzy comparison of fields like name and address that catches typos, nicknames, and formatting differences. Candidates are scored by confidence, so strong matches auto-resolve and borderline ones get review.
Why not just match on exact values?
Because real-world duplicates rarely match exactly - the same person has two emails, a typo in the name, or a company written five ways. Exact matching catches only perfect copies. Combining exact keys with fuzzy comparison catches the messy majority of real duplicates that exact-only matching misses.