Unique Identifiers and Elimination Identifiers

The duplicate identification setup process lets you configure up to three unique identifiers and three elimination identifiers as part of your matching configuration.

Unique identifiers and elimination identifiers are special match tokens that you can use for cases where the value of a single attribute conclusively determines whether a set of records are duplicates, without needing to consider any other attribute values. The meanings of these Unique identifiers and elimination identifiers fields are as follows:

• Unique Identifiers - any records that have the same non-null value for an attribute that has been defined as a unique identifiers are automatically identified as duplicates, regardless of any other attribute values

• Elimination Identifiers - any records that have different non-null values for an attribute that has been defined as an elimination identifiers are never identified as duplicates, regardless of any other attribute values

Both unique identifiers and elimination identifiers only evaluate non-null values, and matching is still possible between records that have a mixture of populated and null elimination identifier or unique identifier values. For example, two records may be identified as duplicates when one record has a value in an elimination identifier field, the other record has no value (null) for the elimination identifier field, and the records also match on other attributes such as name and address.

To better understand the difference between unique identifiers and elimination identifiers, consider the following requirements:

• We will always consider contacts that have the same email address to be duplicates. (This requirement could be handled with a unique identifier field)

• We will never consider contacts that have different email addresses to be duplicates. (This requirement could be handled with an elimination identifier field)

If you choose to use Custom Scoring as part of your duplicate identification configuration, keep the following in mind:

• Records with different non-null elimination identifier values are never identified as duplicates, regardless of any scoring rules. Elimination identifier fields should not be included in custom scoring rules, and adding elimination identifier fields to your custom scoring rules may lead to unpredictable or inconsistent matching results.

• The custom scoring rules for unique identifier fields must always be defined as a single-value exact match rule with a score of 100. The standard matching configuration comes seeded with default custom scoring rules for the three available unique identifier fields, and these default custom scoring rules should not be modified. Unique identifier fields are specifically designed for their function and adding unique identifier fields to any other custom scoring rules may lead to unpredictable or inconsistent matching results.