Data jobs created by field normalization Field Normalization creates four data jobs during the normalization process to collect or change data. All jobs except Pending value collection are executed manually. Data jobs have a roll back feature that allows you to undo normalization at different stages before committing changes that affect the entire database. Pending value collection: This data job is created when a field normalization record is submitted. This job runs automatically to collect all the values from the database for the field to be normalized. This data job does not modify the database. Normal value change: This data job is created when an existing normal value that is used to normalize a field is changed. When the job is run, the platform updates all the normalized fields in the database that use the new normal value. Alias application: This data job is created when an alias is created. When the job is run, all field values in the database containing the alias are normalized. Rule application: This data job is created when a rule is created. When the job is run, all field values in the database that match this rule are normalized. Coalesce to normal: This data job coalesces data from existing records containing a normalized field value into a single record that uses a normal value.