---
title: "Matching Criteria for Dataset Records"
slug: "matching-criteria-for-dataset-records"
updated: 2025-01-31T20:08:43Z
published: 2025-01-31T20:08:43Z
canonical: "knowledge.technolutions.net/matching-criteria-for-dataset-records"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://knowledge.technolutions.net/llms.txt
> Use this file to discover all available pages before exploring further.

# Matching Criteria for Dataset Records

When importing to a Dataset destination (such as Organizations, Funds, or Companies and Foundations) using a form or Upload Dataset, Slate attempts to prevent the creation of duplicate records by finding an existing, matching dataset record. As soon as a dataset record matches a mapped item, Slate considers the record to be found, and it stops evaluating subsequent matching criteria. As such, the order is important to note below.

Forms and imports must have a specific dataset selected, so in addition to the matching criteria listed below, the matching record must also be part of the selected dataset.

## Matching criteria

The items that are used to identify an existing dataset record in Slate appear below in the order provided:

1. **Row Slate GUID Matching Only**
  - This is the 36-character unique identifier of the dataset record.
  - **Matching only:** Used only to find potential matches. Does not update the GUID.
2. **Row Identity Matching Only**
  - This is the integer value of the identity column on the dataset record. It is automatically assigned and cannot be updated.
  - **Matching only:** Used only to find potential matches. Does not update the GUID.
3. **Key**
  - This is the unique key for the dataset record.
  - If the key does not exist, or a match was made from previously evaluated matching criteria, and the key does not match, then the key is updated.
4. **Username**
  - This is the username for the dataset record, which can be used to log in to Portals requiring dataset authentication.
  - If the username does not exist, or a match was made from previously evaluated matching criteria, and the username does not match, then the username is updated.
5. **Dataset-scoped unique custom fields**
  - If the unique field value does not yet exist, or it does not match what was mapped in the source, the value is updated.

> [!WARNING]
> **📝 Note: 64 character limit for unique for merging fields**
> 
> Slate only evaluates the first 64 characters of the values of incoming unique for merging fields. Values greater than 64 characters may result in unexpected behavior.

If a matching record is found, that record is updated.

## If no matching dataset data

If none of these items match, then a new dataset record is created.