Fast Process, Detailed Process, and Custom Settings

The full and incremental processes can run in “fast process” mode, “detailed process” mode, or with custom settings you define. The selected mode determines the settings and filters used for the search process. For more information about specific options and settings, see Full and Incremental Duplicate Search Matching Optionsand Filters.

Fast Process

The Fast process duplicate search is designed for speed and efficiency. In this mode, the search process filters out as many records as it can based on matching parts of the constituent name, email, phone, address, and other features. If two records are matched by this process, there is a high likelihood that they are duplicates. This type of process will generate fewer results than the detailed process, so there is a possibility that some duplicates may be overlooked.

Matching thresholds used:

 •  Required name confidence level to match: Set at 70%

 •  Required address confidence level to match: Set at 70%

 •  Required overall confidence level to match: Set at 70%

Other options used:

 •  Match constituents based on email address

 •  Match constituents based on phone number

 •  Include middle names in comparison

 •  Include maiden name in comparison

 •  Filter individuals by gender

 •  Include Groups

 •  Include Organizations

Additional options not used by the fast process:

 •  Filter individuals by birth date

 •  Pre-filter addresses by state

 •  Pre-filter addresses by zip code (3 digits required for inexact match)

 •  Pre-filter addresses by country

Detailed Process

The Detailed process duplicate search will take longer than the “fast process” but can potentially find more duplicates. This process uses fewer filters to exclude potential matches. It will generate more results than the fast process and may catch some duplicates that the fast process misses. For example, it is much better at finding duplicate constituents with different addresses or phone numbers. Because the detailed process uses fewer filters to exclude records, there is a greater possibility of getting results that are not really duplicates.

Matching thresholds used:

 •  Required name confidence level to match: Set at 70%

 •  Required address confidence level to match: Set at 70%

 •  Required overall confidence level to match: Set at 70%

Other options used:

 •  Match constituents based on email address

 •  Match constituents based on phone number

 •  Include middle names in comparison

 •  Include maiden name in comparison

 •  Filter individuals by gender

 •  Include Groups

 •  Include Organizations

Additional options not used by the detailed process:

 •  Include Inactive

 •  Include Deceased

Custom Settings

Select Custom settings to run a duplicate search based on comparison settings you define manually on the Custom settings tab. You can choose which constituents to compare, which filters to apply, and the minimum confidence level for two records to be considered a match. In most cases, we recommend that you use the fast or detailed process as the selections you make on the Custom settings tab can have lengthy running times.

Tip: If you choose to adjust the match confidence levels on the Custom settings tab, we recommend you start with a low match percentage (the default is 70%). After you run the search process and view the Duplicate Constituents Report, you can compare the results (potential duplicates) to known duplicate records. If all known duplicates are identified, and most potential duplicates are actual duplicates, your match percentage is tuned appropriately. To decrease the similarity required to flag constituents as duplicates, lower the match percentage. To increase the similarity, raise the percentage. By tuning this percentage you minimize the number of duplicate results to manually review. For example, if your initial match percentage is 80% and many of the results are not duplicates, raise the percentage. If all of the results are duplicates, but not all known duplicates are found, lower the percentage.