ADM (Automatic Duplicate Manager) has many fail-safes when it comes to the merge process.  Some are there to prevent what is considered essential or "Not Undoable" data from being merged.  It is common for an Admin to set the predup match score to 100 and expect that anything that has a perfect score to be automatically merged.  This will be the case most of the time, however there are some exceptions:

When ADM skips merge records:
  • Match skipped because both are administrators
  • Match skipped because both have Member IDs
  • Constituent in a private or protected center
  • Match limit exceeded for candidate (matches more than one candidate)
  • One record was updated since being added to predup, but before ADM was run
  • Records that won't be undoable*
*Undoable is defined as having this error when you attempt to merge the records manually or automatically:
User-added image
An example would be if the site settings are set to automatically merge records with a score of 100.  But if the email is set as a score of 0 (ignored), it would mean that 2 records that have the exact same name and address, but different emails, will have a score of 100.  One would assume it would be merged because of the score, but in fact because the emails are different, the system will not assume which one it correct and will require manual review.  The above error message would show on each of those ADM records when attempting to merge.

General ADM Help and Setup:
Automated merge operations occur when only one potential duplicate record meets or exceeds the automatic merge threshold. If more than one potential duplicate constituent records score at or above the high - scoring  threshold, they are not automatically merged and are instead placed in ‘Resolve New Registrations’ for the administrator to reconcile.

The "Predup Score" can be fine tuned and adjusted based on your needs.  Below is a list of what the options are and how the process works:

All matches start with a baseline score of 100. Each rule applies a deduction from the total. Each rule has logic that will deduct the full amount or a percentage of the rule's maximum deduction
**It is important to note that the percentage would be of the amount to take out of the score.  So if the rule is set for 10 points, and it assigns the amount by 10%, it will reduce the score by 1 point.**

To skip a rule, set the value of the max-deduction SDP for that rule to 0.

The checks in each rule below execute in the given order (the rules themselves do not execute in any given order) . If a mismatch is found in a rule, the rule applies the deduction and we move on to the next rule. All rules will execute. The minimum score is 0.

Last Name

SDP: PREDUP_POINTS_LAST_NAME_EXACT (DEFAULT: 40)
  • Exact Match: 0 deduction
  • Last names match with 1 typo; 10% deduction.
  • Last name of one record contains the other ((ex: Smithwick contains Smith); 15% deduction.
  • Last names match for first 7 characters; 12.5% deduction.
  • Last names match for first 5 characters; 20% deduction.
  • Last names match for first 3 characters; 50% deduction.
  • Otherwise, Last names do not match; 100% deduction.

First Name

SDP: PREDUP_POINTS_FIRST_NAME_EXACT (DEFAULT: 30)
  • Either first name is blank; 66% deduction.
  • First names match; no deduction.
  • First names match with 1 typo; 10% deduction.
  • First name of one record contains the other (ex: Joanne contains Anne) ; 15% deduction.
  • First names match for first 5 characters; 20% deduction
  • First names match for first 4 characters; 25% deduction
  • First names match for first 3 characters; 33% deduction
  • First names match for first 2 characters; 66% deduction
  • Otherwise, First names do not match; 100% deduction.

Middle Name

SDP: PREDUP_POINTS_MIDDLE_NAME_EXACT (DEFAULT: 5)
  • Either Middle name is blank; no deduction.
  • First character of middle names do not match; 100% deduction.
  • One field is a single character long and matches other field; no deduction.
  • Middle names match; no deduction.
  • Otherwise, names do not match, 100% deduction

Suffix

SDP: PREDUP_POINTS_SUFFIX_NAME_EXACT (DEFAULT: 15)
  • Either record is missing Suffix; no deduction.
  • Name Suffixes match; no deduction.
  • Otherwise, Name Suffixes do not match; 100% deduction.

Title

SDP: PREDUP_POINTS_TITLE_NAME_EXACT (DEFAULT: 15)
  • Either record is missing Title; no deduction.
  • Name Titles match; no deduction.
  • Otherwise, Name Titles do not match; 100% deduction.

Home Street 1

SDP: PREDUP_POINTS_STREET_EXACT (DEFAULT: 40)
  • Street address is missing from either record; 66% deduction.
  • Streets match; no deduction.
  • Streets have 1 typo: 10% deduction.
  • Streets have 2 typos: 20% deduction.
  • Streets have 3 typos: 30% deduction.
  • One Street contains the other (ex: "123 First Street" contains "First St"), 15% deduction.
  • Streets match for first 10 characters; 10% deduction.
  • Streets match for first 7 characters; 14% deduction.
  • Streets match for first 5 characters; 20% deduction.
  • Streets match for first 4 characters; 33% deduction.
  • Streets match for first 2 characters; 50% deduction.
  • Otherwise, Streets do not match; 100% deduction.

Home City

SDP: PREDUP_POINTS_CITY_EXACT (DEFAULT: 10)
  • City missing from at least one record; 66% deduction.
  • Cities match; no deduction.
  • Cities match with 1 typo; 10% deduction.
  • Otherwise, Cities do not match; 100% deduction.

Home State

SDP: PREDUP_POINTS_STATE_EXACT (DEFAULT: 10)
  • Zip codes match, skipping state comparison; no deduction.
  • State is missing from at either record; 66% deduction.
  • States match; no deduction.
  • Otherwise, States do not match; 100% deduction.

Home Zip

SDP: PREDUP_POINTS_ZIP_EXACT (DEFAULT: 30)
  • Either record is missing zip code; 66% deduction.
  • Zip codes match; no deduction.
  • Zip codes match for first 5 characters; no deduction.
  • Zip codes match for first 4 characters; 25% deduction.
  • Zip codes match for first 2 characters; 66% deduction.
  • Otherwise, Zip codes do not match; 100% deduction.

Home Phone Number

SDP: PREDUP_POINTS_PHONE_EXACT (DEFAULT: 0)
  • Checks Home Street 1, if Streets match, skips the phone comparison; no deduction.
  • Phone is missing from at least one record; no deduction.
  • Phone strings match; no deduction.
  • Numeric-only phones match; no deduction.
  • Otherwise, Phones do not match; 100% deduction.

Primary Email

SDP: PREDUP_POINTS_PRIMARY_EMAIL_EXACT (DEFAULT: 0)
  • Primary email is missing from either record; 66% deduction.
  • Primary emails match; no deduction.
  • Otherwise, Primary emails do not match; 100% deduction.

Household matches

Enabling SDP: PREDUP_MATCH_EXCLUDE_HOUSEHOLD_MEMBERS (DEFAULT: TRUE)

Threshold SDP: PREDUP_MATCH_EXCLUDE_HOUSEHOLD_MEMBERS_THRESHOLD (DEFAULT: 85)

Uses prefix, suffix, and name similarity to determine if the matched records are the same person, or two different members of the same household. If they are household members, returns a full 100 point (not percent) deduction.

Automatic Duplicate Merge Score Threshold

SDP: DUPMGMT_AUTO_DUP_MERGE_SCORE_THRESHOLD (DEFAULT: 100)
(0-100) The minimum score that a match must have before being considered for automatic merge.