Skip to main content
Gainsight Inc.

Unification Overview

This article provides an overview of Gainsight's Unification feature which can unify Company and Person data with various sources and create a unified data repository.


The Unification feature helps you to unify the Company and Person data from various sources within a business line or from multiple business lines and create consolidated Golden Records. You can configure matching rules in the Unification feature to identify duplicates across multiple data sources and resolve issues such as duplicate, incomplete, or inaccurate data, that result in inaccurate reports and data analysis in Gainsight.

If you want to unify data from various sources, please contact the Gainsight Support. The support team contacts the customer to learn about their requirements, and to set up integrations between their applications and Gainsight. All customer data is ingested into Gainsight's staging objects through various data ingestion channels. Administrators can then use these staging objects to configure the Unification project in Gainsight to create a single consolidated Golden Record.

Key Benefits

The following are the key benefits of this feature:

  • You can use the Unification feature:
    • To clean up duplicate data from CRM(s).
    • To avoid duplication of records while unifying data from multiple sources such as:
      • Within a business line, solve data matching problems across functional data sources like sales, finance, support, billing, marketing etc. when there is no common ID propagated in those data sources.
      • Across business lines, solve both matching and duplication problems, which are even more compounded when data from multiple business lines is ingested into the same Gainsight instance.
    • Use both Exact & Fuzzy matching techniques to specify uniqueness criteria.
  • Use the Data Steward feature to view the unified results and manually intervene to resolve duplicate data.
  • With a unified data repository in Gainsight, get a 360 view of your customers across business lines.
  • Automatically model relationships into Gainsight from multiple sources.

Key Terms

The following are key terms that are used in Unification feature:

  • Staging Objects: The Standard and custom objects that need to be unified to build a unified data repository.
  • Match Criteria: The criteria to identify if two records (within the same source or across different sources) are unique or duplicate representations of the same record. The following are the two types of matching criteria:
    • Exact: This option identifies a record as duplicate when there is an exact match between the specified attribute of the records. For example, User’s email Id could be used for exact match as they are mostly considered to be the identical across various sources.
      When comparing two records, if both the records have values for that attribute and they match exactly, the records are considered duplicates.
      If the values do not match, the records are considered unique. In case, either of the records do not have values for that attribute, the next rule is triggered to determine whether these records are duplicates.
    • Fuzzy: This option identifies a record as duplicate when the two records have similar but not the exact same values between the specific attributes of the records. A match is considered when the matching score for the values is equal or greater than the threshold set in the rule. For example, Company Name can be used for fuzzy match as it might differ slightly in various sources like IBM or IBM Inc., Example or Exampel, etc.
      When comparing two records, if both the records have values for that attribute and they match fuzzily, the records are considered duplicates.
      If the values do not match fuzzily or if either of the records do not have values, the next rule is evaluated to determine whether these records are duplicates.

unnamed .png

  • Cluster: Group of records that have been identified as duplicates across various sources
  • Golden Records: The consolidated record created as a result of unification of contributing records of the cluster. Unification picks the values based on the preferred sources for each attribute. Golden records are inserted into Gainsight as the final step of the Unification process.
  • Data Steward: This feature helps you to supervise Unification results and manually intervene if needed. You can also perform the following manual tasks:
    • Explore more matches for existing clusters.
    • Move records out of existing clusters to other existing clusters or a new cluster.
    • Trigger a manual merge of Company records within Gainsight.
    • Manually edit the attributes of Golden records.

Architectural Diagram

The following diagram shows how the Unification feature builds a unified data repository from CRMs and other data sources:



Step 1: Using Rules & Connectors, data can be ingested directly into Standard objects or Custom objects within Gainsight. Both these objects together are called Staging objects.

Step 2: Once the staging objects are ready, admins can perform the following activities:

  1. Add the required staging objects to the project.
  2. Map the Staging objects to the Company/Person models
  3. Use identification criteria to group duplicate data into clusters

Step 3: After the Project configuration is completed, Unification feature builds a consolidated Golden record for each cluster and Golden records are promoted to standard objects.

Step 4: Data steward helps in supervising automated unification or for manual intervention.

Data Steward

When using the Unification feature to combine data from multiple sources, it's critical to make sure the final data is free of duplicate, incomplete, or inaccurate records. The matching criteria set by the Admin when creating the project helps to solve this problem in a scalable manner.

The Data Steward feature assists in manually resolving duplicate data that the system was unable to resolve. It also helps in the undoing of the unification feature's changes in Gainsight.

Additional Resources

For more information on Unification feature, refer to the following articles:

  • Was this article helpful?