Device Deduplication > Detection Settings Reference

The Deduplicator supports three pre-set detection options: Broad, Standard, and Precise. Each pre-set defines a fixed combination of global similarity thresholds, field weight values, and field matching thresholds. Selecting a pre-set applies all associated values at once. This reference lists the exact values used by each pre-set so you can assess and compare detection behavior without switching to Custom mode.

When Custom is selected, all values listed here serve as a reference for defining your own configuration.

Global Similarity Threshold

The global similarity threshold is the minimum combined similarity score that two records must reach before they are considered duplicates. A lower threshold produces more matches (broader detection); a higher threshold requires closer similarity (more precise detection).

Pre-set Option	Global Similarity Threshold	Strong Identifier Match
Broad	40	No
Standard	65	No
Precise	95	Yes

Strong Identifier Match — When enabled (Precise only), the Deduplicator requires that at least one strong identifier field (such as UUID or Serial Number) matches exactly before records can be merged. This prevents false positives caused by coincidental similarity in soft fields.

Field Weights

Field weights determine how much each field contributes to the overall similarity score between two records. A higher weight means the field has more influence on whether two records are considered duplicates. Field weight values are the same across all three pre-set options. They can only be adjusted in Custom mode.

Devices

Field	Weight
UUID	30
Serial Number	25
BIOS	15
MAC Address	10
Device ID	5
FQDN	5
Host Name	5
Domain	3
IP Address	2

Software

Field	Weight
Product	40
Publisher	30
Product Version	20
Edition	5
Language Code	5

Users

Field	Weight
Email	65
User Principal Name	65
Login	35
User Key	35
Import ID	20
First Name	10
Last Name	10

Field Matching Thresholds

The field matching threshold is the minimum similarity score that two field values must reach before the field contributes positively to the overall similarity calculation. Values below the threshold are treated as non-matching for that field. Unlike field weights, the matching thresholds differ between pre-set options — Broad uses lower thresholds to accept more approximate matches, while Precise requires near-identical values.

Devices

Field	Broad	Standard	Precise
UUID	60	100	100
Serial Number	60	100	100
BIOS	60	100	100
MAC Address	40	80	95
Device ID	50	100	100
FQDN	40	80	95
Host Name	40	80	95
Domain	50	90	100
IP Address	40	70	90

Software

Field	Broad	Standard	Precise
Product	60	100	100
Publisher	60	100	100
Product Version	60	100	100
Edition	50	90	100
Language Code	50	90	100

Users

Field	Broad	Standard	Precise
Email	60	100	100
User Principal Name	60	100	100
Login	60	100	100
User Key	60	100	100
Import ID	60	100	100
First Name	60	85	100
Last Name	60	85	100