Custom Duplicate Groups is an alternative, rule-based way to catch duplicate invoices. It is designed to guarantee that certain kinds of duplicates are always raised, at the expense of a higher false positive rate đ§©
Contents:
1. How Custom Duplicate Groups works
In addition to our standard duplicate detection system (see How we detect duplicate invoices), we also offer an alternative, rule-based approach called Custom Duplicate Groups. This can guarantee that certain kinds of duplicates are raised, at the expense of an increased false positive rate.
The process works in three steps:
Retrieve all invoices in the database.
Use a set of rules to put the invoices into groups (see the grouping rules below).
Within each group, select the two most recent invoices and raise them as a duplicate pair in the platform.
That final step keeps us from raising an overwhelming number of duplicates from a small number of large groups. For example, a group of 100 invoices could naively produce almost 5,000 possible duplicate pairs - by only raising the two most recent invoices per group, we keep the results manageable and relevant.
Once raised, a Custom Duplicate Groups pair appears in the Duplicate Invoices tab just like any other duplicate pair, ready for you to classify and resolve. To learn how, check Resolve Duplicate Pairings.
2. Grouping rules available
High Value
Groups invoices by gross amount and vendor. Only raises duplicate pairs where both invoices are above a configurable threshold value.
High Risk Vendors
Groups invoices by gross amount and vendor. Only raises duplicate pairs where both invoices are from one of a user-specified list of "high risk" vendors.
Optionally, can also be restricted to only raise pairs where both invoices are above a threshold value.
Same number, vendor name
Groups invoices by invoice number and vendor name.
Optionally, can be restricted to only raise pairs where the invoices have different amounts, and/or where both invoices are above a threshold value.
Same cleaned number, date, vendor name
Groups invoices by cleaned invoice number (number with all non-numerical characters removed), date and vendor name.
Optionally, can be restricted to only raise pairs where the invoices have different amounts, and/or where both invoices are above a threshold value.
Same number, date, vendor name
Groups invoices by invoice number, date and vendor name.
Optionally, can be restricted to only raise pairs where the invoices have different amounts, and/or where both invoices are above a threshold value.
3. Turn this on for your organisation
By default, no Custom Duplicate Groups rules are set up for your account. Because this approach trades off a higher false positive rate for the certainty of catching specific kinds of duplicates, we set it up together with you rather than turning it on automatically.
If you'd like to use one or more of the grouping rules above, get in touch with your dedicated Account Manager to discuss which rules and thresholds make sense for your data.
Want to learn more about how duplicates are found and managed? Check Behind-the-scenes: how we detect duplicate invoices, Duplicate Invoice Configurations, or Overview of Duplicate Invoices Tab.