Understanding Entity Identifiers
When you configure a data preparation process for an Identity Engine workflow, you must configure a Primary ID entity identifier for your data source. You can also optionally map additional identifiers, such as Customer ID and Household ID, in the Identifier entity.
These additional identifiers are not required for resolution to work, but mapping them is strongly recommended. They help you:
Preserve your own business keys across systems and sources
Understand and tune consolidation behavior
Monitor graph health and trends in dashboards over time
This topic explains how Primary ID, Customer ID, and Household ID differ, how Identity Engine uses them along with Enterprise ID, and best practices for configuring them in a data preparation process.
Configured Entity Identifiers and Enterprise IDs
The Primary ID, Customer ID, and Household ID are entity identifiers that you configure in a data preparation process. An Enterprise ID is a unique identifier that Identity Engine generates during resolution and maintains to represent a person, household, business, or other consolidated entity that is relevant to your organization.
Identity Engine often uses an Enterprise ID alongside a Primary ID, Customer ID, and/or Household ID for reporting and metrics. The Graph Build Overview dashboard and Workflow Trends dashboard use these identifiers to display charts with useful metrics and trends to help you improve the efficiency of your workflow processes and the quality of your identity graph.
For more information on the relationship between configured entity identifiers and Enterprise IDs, see “Graph Build Overview Dashboard” and “Workflow Trends Dashboard”.
Primary ID, Customer ID, and Household ID
In the “Entity Mapping” step of a data preparation process, the Identifier entity exposes three key fields:
Primary ID: A required unique identifier for each unique row of a data source. Identity Engine uses the Primary ID for data quality gates and resolution behavior. It must be unique, consistent, and persistent for each entity within and across files for the same source. For more information, see “Configure a Primary ID”.
Customer ID: A business-level identifier that can appear in multiple rows and multiple sources, such as an account number or CRM ID.
Household ID: A group-level identifier that represents a household across one or more sources and can appear in multiple rows and multiple sources.
Note
Because Customer ID and Household ID may have duplicates, they may not be appropriate as Primary IDs in most data sets. However, they are important identifiers that Identity Engine tracks and surfaces for analysis and monitoring.
Using a Customer ID
Configuring a Customer ID in a data preparation process gives Identity Engine an explicit representation of an account or customer key in your data source. This supports monitoring, troubleshooting, and downstream usage in the following ways:
On the Graph Build Overview dashboard, the Consolidation Rate chart compares:
Input Records: Total number of unique rows in the input data
Customer IDs: Number of unique values mapped as Customer ID in your data preparation processes
Enterprise IDs: Number of unique Enterprise IDs in the graph
If Customer ID is not configured, the chart will use the Primary ID. Seeing how records consolidate from Input Records to Customer IDs to Enterprise IDs helps you:
Identify over consolidation or under consolidation, for example Enterprise IDs nearly equal to Customer IDs versus significantly lower than Customer IDs
Diagnose whether unexpected changes are driven by source data changes or by resolution rules
On the Workflow Trends dashboard, the Resolution Trends chart shows trends over time in:
Customer IDs
Enterprise IDs
Total input records
Comparing how Customer IDs and Enterprise IDs move relative to each other across runs helps you identify when:
Resolution rules changed and had an unexpected impact
New or modified data sources changed consolidation behavior
Customer keys are being unintentionally split or over merged
Without a Customer ID mapped, these trends are less informative, because you lose the intermediate business-level identifier between total input rows and Enterprise IDs.
Downstream Usage
Many downstream usage scenarios handle data in terms of customer accounts rather than Primary IDs. Configuring a Customer ID ensures that:
Exports can include both Enterprise ID and the original Customer ID, so users can join results back to CRM, billing, or other internal systems.
Dashboards and analysis can be interpreted using a familiar identifier, Customer ID, alongside Enterprise IDs and input record counts.
For more information on configuring a Customer ID, see “Configure a Primary ID”.
Using a Household ID
If your organization models relationships at the household level, configuring a Household ID gives Identity Engine a useful identifier for that grouping. A Household ID identifier can be common across several sources and have duplicates within a source. Identity Engine can also use a Household ID you configure to generate Household Enterprise IDs, which are Enterprise IDs that represent households rather than individuals.
By configuring a Household ID, you:
Preserve your own notion of a household, for example an internal household key from your CRM
Enable metrics such as Household Enterprise IDs on dashboards to be interpreted in the context of your existing household definitions
Support downstream reporting and activation use cases that require both person level and household level views
If you do not map a Household ID, Identity Engine can still infer household level groupings based on rules and graph behavior. However, you lose the direct connection back to your own household key.
For more information on configuring a Customer ID, see “Configure a Primary ID”.
Best Practices for Using Entity Identifiers
Follow these best practices when working with Primary ID, Customer ID, and Household ID identifiers:
Configuring a Primary ID is required.
You should only use the “Automatically Configure Primary ID” option when neither a single field nor a combination of multiple fields can be configured to result in a unique and persistent Primary ID for each entity in your dataset.
Configure a Customer ID whenever you have a suitable field to provide an account level view across sources and time.
Use your existing customer key if one exists, and keep it stable over time so trends and metrics remain interpretable.
Configure a Household ID if household-level analysis, reporting, or activation is important for your use cases.
Use your existing household key if one exists, and keep it stable over time so trends and metrics remain interpretable.
Keep the meaning of each identifier stable. Do not repurpose a Customer ID or Household ID column for a different business concept.
If your business changes how these IDs are generated, coordinate that change with your Identity Engine configuration.
Review Consolidation Rate and Resolution Trends after changes to identifier mapping or resolution rules.
Investigate unexpected shifts between Customer IDs and Enterprise IDs, or large changes in household related metrics.
By combining a unique, consistent, and persistent Primary ID with the additional Customer ID and Household ID entity identifiers, you give Identity Engine the information it needs to build a reliable first party graph while keeping results interpretable and actionable.