Deduplication rules

Overview

Sometimes, the same individual might have multiple accounts in your ticketing system. Ordinarily, this would lead to multiple CRM contacts for the same person. KORE Ticketing solves this problem by associating duplicate ticketing accounts with a single CRM contact. This page explains the logic we use for deduplication.

Identifying duplicates

CRM contacts

When KORE Ticketing encounters a ticketing account for the first time (that is, the ticketing account isn’t listed in the mapping table), it searches for a matching CRM contact. It only creates a new CRM contact if no match is found. A ticketing account matches a CRM contact if one of these rules is satisfied:

• Email address (exact)
• OR all of these:
  • Last name (exact)
  • First name
  • Company name
  • Address
  • Postal code
• OR all of these:
  • Last name (exact)
  • First name
  • Company name
  • Business phone

None of these rules are case-sensitive, but email addresses and last (family) names must otherwise match exactly. We use fuzzy matching (described below) for the other fields.

If a field (excluding first or last name) is blank in both records, this will be considered a match. However, at least one pair of the listed fields (excluding first or last name) must be filled out for the second or third rule to be satisfied. For example, consider these two records:

The first rule isn’t satisfied because of a typo in the second record’s email field—it’s not an exact match. The second rule isn’t satisfied because the postal codes don’t match, even though the other criteria are met.

The third rule is satisfied:

• Last name (exact)
• First name (fuzzy)
• Company name (blank)
• Business phone (fuzzy)
• At least one match (other than name) isn’t blank

If the business phone fields had instead been blank in both records, the third rule would not be satisfied—there would no longer be a non-blank match that isn’t a name field.

CRM accounts

If a ticketing account includes a company name, the corresponding CRM contact will be connected to a CRM account representing that company. To prevent creating multiple CRM accounts for the same company, KORE Ticketing first searches for an existing CRM account. If the company name from the ticketing account fuzzy matches a CRM account’s company name, the existing CRM account will be used.

Fuzzy matching

Fuzzy matching is a technique used to identify information that may not be written in the same way, but which likely still refers to the same thing. Because fuzzy matching is inexact, it’s important to compare multiple fields before labeling two records as duplicates.

First name

Many names are commonly shortened. For example, “Robert” may use the name “Bob” or “Jonathan” might use “Jon”. We use an extensive list of such aliases to make fuzzy matches. We also consider common spelling variations for fuzzy matching. For example, “Muhammad” and “Mohammad” would match, as would “Jeff” and “Geoff”. However, other differences like unexpected typos are not matched. For example, “Velma” wouldn’t match “Vekma”.

Company name

Before comparing two company names, we remove common prefixes and postfixes. We also remove spaces, remove or alter punctuation marks, and change all letters to lower case. (This is only for comparison purposes and does not replace the company name in your CRM.)

For example, we would convert “Fed Ex Inc” to “fedex” without the spaces or postfix. If another record contains the company name “Fed-ex Ltd”, we would likewise convert it to “fedex” which fuzzy matches “Fed Ex Inc”.

Address

We compare the portion of an address which comes before the first space. For example, “129 First St” and “129 First Street” match since they both begin with “129”. Either one would also match “129 Second Ave”, but a single fuzzy match alone is insufficient to identify two CRM contacts as duplicates.

Postal code

For US addresses, we remove the “plus four” digits (if present) and only compare the base ZIP codes. For example, “00210-1234” would fuzzy match both “00210” and “00210-5678”.

Phone number

We normalize phone numbers into a standard format prior to comparison. For example, 800-555-0109 would match +1 (800) 555-0109. If a phone number includes an extension, it is preserved and must be matched. For example, 800-555-0109 Ext. 287 would not match 800-555-0109 (without an extension) nor 800-555-0109 Ext. 176.