How duplicate matching works
The signals DeDupe compares to decide two records are the same.
DeDupe compares records across several signals and combines them into a single confidence score. No single field decides a match on its own.
Matching rules
Your workspace's matching rules decide what counts as a duplicate. Each rule looks at one or more signals:
Companies — similar domain (tolerant of typos like acme.com vs
acme.co) and similar company name ("Acme Inc" vs "Acme, Inc.").
People — similar email, same phone number (normalised to a common format), same LinkedIn profile, same Twitter handle, similar name, same name + company, and same name + email domain.
A rule fires when all of its conditions match, and a pair is flagged if any enabled rule fires. Admins can switch built-in rules on or off, set how strict each one is, add custom rules ("same LinkedIn and same phone"), and add field clean-up (ignore middle names, treat "Bob" and "Robert" as the same).
Why this matched
Open any pair in the review queue to see Why this matched: every rule that fired, the values it compared, how similar they were, and how the confidence score was put together.
Why some obvious duplicates rank lower
If two records share only a name and nothing else, DeDupe stays cautious — common company and person names produce false matches. Filling in more complete data (domains, emails, phone numbers) improves match quality over time.
Next
See Understanding confidence scores for how the signals combine into the High / Medium / Low bands you see in the queue.