Identify How HubSpot Duplicates Were Created and Prevent Them Using Insycle

There are a few nuances with the Original Source fields that are helpful to understand upfront.

Field Names May Vary By Object Type

The record source field names can be different among object types, but the internal name remains consistent and serves the same purpose across all object types.

Insycle Shows IDs, While HubSpot Displays Readable Names

When viewing the Original Source [Data] 2 field in Insycle's Step 2 preview or exported CSV, you'll see numeric IDs such as "581872", "68778994", or "userId:45798207". These same fields display more readable information in HubSpot, such as:

• Integration names (e.g., "Insycle Data Management", "Salesforce")
• CSV filenames (e.g., "eways-conference-contacts.csv")
• User names and emails (e.g., "Lana Sothern (lana@vurtego.org)")
• Workflow names (e.g., "MQL to SQL Conversion")

The numeric IDs in Insycle correspond directly to these readable names in HubSpot. To determine what a specific ID represents, you can look up the corresponding record in HubSpot to view the full details.

Focus on Original Source [Data] 1 For Pattern Identification

The top-level Original Source field shows very broad categories like "Offline Sources" or "Direct Traffic," which aren't specific enough to identify duplicate creation patterns. The most useful field for analysis is Original Source [Data] 1, where you'll see specific source types such as:

• IMPORT - Records created via CSV import
• INTEGRATION - Records created via third-party integrations
• CRM_UI - Records manually created in HubSpot
• Form URLs - Specific forms that created records
• EXTENSION - Records created by the sales extension, Chrome extension, etc.

When analyzing duplicate patterns, pay close attention to the Original Source [Data] 1 field to identify which creation method is responsible for the duplicates.

Learn more about HubSpot’s record source fields.

The first method involves adding Original Source fields directly to your matching rules in Step 1. When you include these fields as matching criteria, Insycle will only identify records as duplicates if they match on all specified fields, including Original Source values.

For example, if you configure your matching rules to include Email, First Name, Last Name, and Original Source, Insycle will only count contacts that have matching values for all four fields as duplicates. This means two contacts with the same email and name but different Original Source values would not be grouped as duplicates.

When to use this method:

• You want to find duplicates that came from the same specific source
• You're investigating a known duplicate creation issue from a particular channel
• You want to segment your deduplication efforts by source type

This approach is most valuable when you already know which source is causing problems and want to isolate those specific duplicates for targeted cleanup.

How to Add Original Source Fields as Matching Criteria

Navigate to Data Management > Merge Duplicates and select the record type.

In Step 1, click +FIELD to add a new matching row. Select Original Source from the field dropdown, and set the Comparison Rule to Exact Match. 

Optionally, add Original Source Drill-Down 1 and/or Original Source Drill-Down 2 as additional matching criteria.

Combine with other unique identifying fields (e.g., email, Name, company domain, etc.).

Click FIND to analyze for duplicates.

Use Case Examples

Example 1: Finding Import Duplicates

• Scenario: You suspect a recent import created duplicates
• Configuration: Match on Name and Email; Filter on Original Source Drill-Down 1 = "Import"
• Result: Identifies only duplicates that came from imports, allowing focused cleanup

Example 2: Integration-Specific Duplicates

• Scenario: Your Salesforce sync is creating duplicate contacts
• Configuration: Match on Email and Original Source; Filter on Original Source = "Integration" + Original Source Drill-Down 1 = the Salesforce sync ID
• Result: Surfaces duplicates specifically from the Salesforce integration

Adding the Original Source fields to your column layout in Step 2 makes the data visible in the duplicate preview table and in exported CSV reports, without restricting which duplicates are identified.

Why this method is recommended:

• You discover patterns across all your duplicates, not just one source
• You don't need to know the source beforehand—the data reveals it to you
• You can identify which sources are creating the most duplicates
• It provides context that helps you make smarter master record selections
• It creates a historical record when you export the CSV before merging

This method takes a diagnostic approach. Rather than focusing your analysis on one source at a time, you can see the complete picture of where your duplicates are coming from and identify patterns you might not have expected.

Add Original Source Fields to Your Template Layout

Navigate to Data Management > Merge Duplicates and select your record type (contacts, companies, or deals). Then, configure Step 1 with your standard matching rules (e.g., Email, Name, etc.) and click FIND.

In Step 2, click the checkbox by a duplicate group to expand the results and review the column headers in the duplicate preview table.

Click the gear icon on the right side of the Step 2 header.

Search for the following fields in the right column:

• Original Source
• Original Source Drill-Down 1
• Original Source Drill-Down 2

Drag each one left to the Visible Fields column, then drag-and-drop them to arrange the columns (typically below basic identifying fields).

With this approach, you can identify the origin of each record in every duplicate group, revealing patterns and insights into how your duplicates are being created.

Reading and Interpreting the Step 2 Preview

Once you've added the Original Source fields to your layout, you'll see them in the duplicate preview table. Here's what to look for:

• Identical Original Source values: All records in a duplicate group came from the same channel (e.g., all from same import)
• Mixed Original Source values: Records came from different channels (e.g., one from Form, one from Import)
• Integration IDs: Drill-Down 2 often shows specific integration IDs or import batch numbers
• Volume patterns: Many duplicate groups showing the same source indicate a systematic issue

In the above example, all four duplicate records show:

• Original Source = "Offline Sources"
• Original Source Drill-Down 1 = "INTEGRATION"
• Original Source Drill-Down 2 = "68300453" (integration ID)

This indicates that all these duplicates originated from the same integration (ID 68300453), suggesting a systematic problem with that integration's deduplication logic.

These two methods aren't mutually exclusive. You can start with Method 2 to analyze your overall duplicate landscape and identify patterns. Once you discover what specific source is creating many duplicates, you may then use Method 1 to isolate and address those source-specific duplicates separately.

For example, you might first run an analysis with Original Source fields visible in Step 2 and discover that 40% of your duplicates are coming from a particular integration. You could then create a second template that uses Method 1—adding Original Source as a matching criterion—to focus exclusively on cleaning up those integration-created duplicates while you work on fixing the root cause.

When you see forms in the Original Source Data, it means form submissions on your website or landing pages are creating duplicate records. This is a common duplicate pattern and usually indicates a form configuration issue.

What you'll see in the data:

• Original Source = "Offline Sources" or "Direct Traffic"
• Original Source Data 1 = URL of the form or page where the form was submitted
• Original Source Data 2 = form ID (displayed as a number in Insycle, may show form name in HubSpot)

Common Form-Related Duplicate Scenarios

Multiple Form Submissions

A contact submits forms multiple times—either the same form repeatedly or different forms across your site—and HubSpot creates a new contact record with each submission rather than updating the existing record.

Pattern clue: You see duplicate groups where all records show the same email address but different forms and dates

This indicates your forms aren't recognizing and updating existing contacts across your form portfolio. it's a clear signal to review your form's settings and update them to prevent creating duplicates by default, instead of creating new contacts.

Integration and Tracking Issues

Form submissions from different devices, browsers, or third-party platforms create separate records rather than recognizing the returning visitor.

Pattern clue: You see duplicates where the same person submitted forms from different contexts:

• Same form submitted from work computer and mobile device
• Forms hosted on external platforms (WordPress, Unbounce, landing page builders) that don't check for existing contacts
• Submissions in incognito mode or with cookies disabled

This pattern appears when HubSpot's tracking cookie isn't working properly, or when third-party form integrations lack proper deduplication logic.

Remediation actions needed:

• Ensure all forms are set to "Update existing contact" instead of "Always create [contact]" 
• Review tracking cookie configuration on your site
• Examine third-party form builders or landing page tools and verify that the integration includes proper deduplication logic
• Consider progressive profiling for returning visitors

When you see duplicates with "Import" as the Original Source Data 1, it means that CSV imports into HubSpot are causing duplicate records. Import-related duplicates can involve large batches of records added to your CRM at once, and issues during import can create dozens or even hundreds of duplicates in a single operation.

What you'll see in the data:

• Original Source = "Offline Sources"
• Original Source Data 1 = "IMPORT"
• Original Source Data 2 = The same import ID number (in Insycle) or CSV filename (in HubSpot, e.g., "conference-contacts.csv")

Common Import-Related Duplicate Scenarios

Single Import Creating Multiple Records for One Contact

A single import file may include the same contact multiple times—either due to duplicate entries in the CSV or because the same email, name, or other identifier appears with minor differences (such as different capitalization, extra spaces, etc.).

You see a duplicate group where all records show the same source data, date, and time, and the same contact appears multiple times with identical or nearly identical data.

This happens when CSVs aren't cleaned and deduplicated before uploading to HubSpot. All records share the same import ID and create date, indicating they came from the same file at the same time.

Repeated Imports of the Same Data

The same list of contacts is imported multiple times over weeks or months, with each import creating new records rather than updating existing ones.

Pattern clue: You see a duplicate group with three records for the same person, all with different create dates.

This pattern usually occurs with recurring imports (monthly partner lists, quarterly customer exports, weekly lead lists from external systems) that lack deduplication logic.

Import Adding Records That Already Exist from Other Sources

Team members import data without coordinating with each other or verifying existing records, or imports add records for people who already exist in other sources, such as form submissions or integrations.

Pattern clue: You see a duplicate group with two records with the same name and or email, but different source data and creation dates.

The contact signed up via your newsletter form in January and attended your trade show in March. When you imported the trade show attendee list, HubSpot created a duplicate instead of updating the existing record. This indicates a lack of standardized import procedures across your team.

Remediation actions needed:

• Always clean and deduplicate CSV files before importing to HubSpot
• Use Insycle's Magical Import module, which includes built-in deduplication
• Review your import mapping settings to ensure unique identifiers (like email) are properly configured
• Establish a standard import procedure for your team
• Merge the existing duplicates after implementing preventive measures

When you see "Integration" in the Original Source Data, it means third-party integrations syncing data with HubSpot are creating duplicate records. Integration-related duplicates are particularly problematic because they're automated—once an integration starts creating them, it can generate many quickly without manual intervention.

What you'll see in the data:

• Original Source = "Offline Sources"
• Original Source Data 1 = "INTEGRATION"
• Original Source Data 2 = integration ID number (in Insycle) or integration name (in HubSpot, e.g., "Zapier", "Salesforce", "Zoom")

Common Integration-Related Duplicate Scenarios

Bidirectional Sync Creating Duplicates

When two systems sync data back and forth (such as HubSpot and Salesforce), a record can be created in one system, synchronized to the other, and then, due to mismatched field mappings or sync rules, synced back as a "new" record, resulting in a duplicate.

Pattern clue: You see a duplicate group where at least one record contains integration info in the Original Source fields, and the created date of that record is after an existing record. You see multiple duplicate groups with this, all with the same integration ID.

For example, this might happen when a contact submits your form in HubSpot, which then syncs to Salesforce, but is later synced back to HubSpot as a new record because the field mapping didn't correctly recognize it as the same contact. A third record might have been created when the Salesforce record was updated.

Multiple Integrations Creating Records for the Same Entity

There are different integrations, and they're not aware of each other's activities. Each can create separate records for the same person or company because they don't check whether a record already exists in HubSpot.

Pattern clue: All records in a duplicate group share the same name and other identifying values, but they were created by different integrations at various times. 

This suggests that each integration is operating independently, without verifying whether existing records exist.

Missing or Incorrect Unique Identifiers

The integration's field mapping is incorrect, causing the system to fail to recognize existing records as matches. This can happen when the unique identifier field (such as an email address, domain + name, or record ID) isn't properly mapped or mapped to the wrong field, or when the selected unique identifier isn't appropriate for the situation.

Pattern clue: At least one record across several duplicate groups shares integration details in the Original Source fields.

This pattern indicates that during a single sync operation, the integration created new records instead of updating existing ones because it was unable to match them correctly.

Remediation actions needed:

1. Identify which integration corresponds to the ID number in the original source field from your HubSpot data
2. Pause or disable that integration to prevent further duplicates
3. Contact the integration provider or review the integration's configuration
4. Check the integration's field mappings, particularly the unique identifier field
5. Test the integration with a small data set before re-enabling
6. Merge the existing duplicates only after confirming the integration issue is fixed 

Mixed creation method duplicates are fundamentally different from the patterns discussed earlier. Instead of indicating a technical configuration issue with a single channel, these duplicates usually represent a person or company legitimately engaging with your business through multiple touchpoints—each interaction generating a separate record because your systems aren't unified in their deduplication approach.

What you'll see in the data:

• Records within a duplicate group show different Original Source Data values
• Different original source data values indicate distinct channels. For example: One record = "FORMS", another = "IMPORT", a third = "INTEGRATION"
• Create Dates typically spread across different days, weeks, or months

Common Mixed Creation Method Scenarios

Multi-Touch Customer Journey Creating Separate Records

A prospect interacts with your business through various channels over time, and each channel creates a new record instead of updating the existing one.

Example: You see a duplicate group with three records with similar identifying information. Perhaps a prospect downloaded an eBook in January (creating the first record via form), then a sales rep manually added them to Salesforce in February after a cold outreach call (syncing to HubSpot as a second record), and finally attended your webinar in March (imported as a third record). 

This indicates that each system or process operated independently, without checking existing records.

Personal vs. Professional Email Addresses

The same person uses different email addresses (personal vs. work) when interacting with your business, and your systems treat these as entirely separate individuals.

Pattern clue: Duplicate groups have similar names, but the email addresses may use one professional and one person (i.e., @gmail.com).

For example, John initially engaged with your sales team using his work email, creating a record through your Salesforce integration. Later, he registered for a webinar using his personal email address (perhaps from home after hours), creating a second record. Because the email addresses differ, automated deduplication didn't catch this.

Note: In this case, you should carefully evaluate whether these need to be merged. Sometimes separate email addresses legitimately represent a desire for separate communication streams (personal vs professional), while other times they're truly duplicates that should be consolidated.

Cross-Team and Cross-Platform Issues

Different departments, team members, or lead-generation platforms create records for the same person using their preferred tools or workflows, without coordination.

Pattern clue: You see a duplicate group with similar identifying info, but each was created by a different team member or process.

For example, a sales rep manually added a contact after meeting someone at a conference on June 1. Two days later, that same person submitted a contact form on June 3. Two weeks after that, on June 15, marketing imported a list from an event that included this person. Each action created a new record because the different creation methods didn't cross-reference each other.

These patterns show siloed team workflows or multi-platform lead generation without unified deduplication.

Remediation actions needed:

• Ensure you have processes in place to check for cross-channel (forms, imports, integrations, manual entry) duplicates.
• Configure your forms to update existing contacts
• Use Insycle's Magical Import for future attendee list imports, including deduplication
• Review your integration sync settings to check for existing records
• Consider implementing a unified deduplication strategy across all channels

Custom code or applications using HubSpot's API may generate duplicates if they fail to check for existing records before creation.

What you'll see in the data:

• Original Source Data fields may show "API", specific API application names, API keys, or remain blank
• Create Dates may show clustered time patterns

Common API Duplication Creation Scenarios

Custom Application Creating Records Without Duplicate Checking

A custom-built application, internal tool, or third-party app uses HubSpot's API to create records but doesn't include logic to search for existing records first.

Pattern clue: You see duplicate groups where Original Source Data contains technical identifiers (API keys, app IDs) for some or all of the records. There are consistent timing patterns (hourly, daily, triggered by specific events).

Example: A prospect submitted a demo request form, creating a contact. Your custom lead routing application then processed the form submission and attempted to create a contact via API without first checking if the form submission had already created one.

API Integration with Incomplete Error Handling

A custom API integration attempts to update existing records, but when it encounters an error or can't find a match, it defaults to creating a new record instead of failing gracefully or alerting someone to the issue.

Pattern clue: Each duplicate group contains records that have the same API or custom application details in the Original Source fields. Error codes are present.

For example, your custom application tried to update existing records via API, but due to field mapping issues, permission problems, or data format mismatches, the update failed, and the application created new records instead.

Remediation actions needed:

1. Review the custom application code and confirm it includes logic to check for existing records before creating new ones
2. Implement proper error handling if an update operation fails
3. Test application functionality with one record or in a sandbox environment
4. Consider using HubSpot's bulk API endpoints, which include better duplicate handling

Manually created duplicate records entered manually by team members directly in HubSpot without checking for existing records before creation.

What you'll see in the data:

• Original Source = "Offline Sources"
• Original Source Data shows user information
• Record Owner field often populated with the person who manually created it
• Scattered, irregular timing (whenever team members happen to create records)
• Usually affects smaller volumes but can be persistent over time

Common Manual Creation Duplicate Scenarios

Sales team members manually create contact or company records in HubSpot without first searching to see if those records already exist in the database, or different team members encounter the same person at different times, and each manually creates a record.

Remediation actions needed:

1. Establish processes or guidelines for when and how team members should manually create records
2. Train the team to always search by multiple data points before manually creating records, such as:
   • Email address plus name for contacts
   • Company name plus domain or address for companies
   • And to search for variation or typos (use Similar match instead of Exact match, or vary the matching criteria)
3. Review whether the team should be creating contacts manually at all, or if there's a better automated process
4. Check if this pattern appears frequently with specific team members who might need additional training

  Note that workflow source data is captured in the "Record Source" fields, rather than "Origin Source."

Workflows can use the Create record action to automatically create new contacts, companies, deals (and other objects). Workflow-created duplicates can occur when your criteria aren’t strict enough, enrollment triggers re-enroll records, or the workflow doesn't first check whether records already exist. 

Because workflows are automated and can process many records quickly, a misconfigured workflow can create a large number of duplicates in a short period.

What you'll see in the data:

• Record Source = "AUTOMATION_PLATFORM" in Insycle, or "Workflow" in HubSpot
• Record Source Detail 1 = workflow name in Insycle and HubSpot

Workflow-Related Duplicate Scenarios

Workflow Re-Creating Records After Deletion

A workflow creates a record based on certain conditions. If the record is later deleted, but the workflow trigger conditions are still met, the workflow will create a new record.

Pattern clue: You see multiple records for the same entity, showing the same workflow info in the Record Source data, with different dates that match the workflow intervals.

This pattern indicates the company was deleted and recreated multiple times, possibly due to repeated manual deletions or a workflow conflict (one creating, another deleting).

Workflow Looking But Not Finding Records Due to Data Inconsistencies

If a workflow creates records based on field values from other records but doesn't account for data inconsistencies (such as capitalization, extra spaces, or punctuation), different email or no email (for contacts), or a different/no domain (for companies), it may result in additional, possibly duplicate records.

Pattern clue: You see similar, but not identical, duplicate records. The creation date of the record with a workflow record source is later than that of the other duplicate. 

A workflow designed to automatically create company or deal records based on contact information creates duplicates when similar but not identical data is present.

The workflow uses contact ‘Company’ fields to create company records; however, if the company names are formatted differently from the existing record, it creates a duplicate instead of associating the contacts with the existing companies.

Same Record Type Creation Triggers Enrollment-Create Cycle

When a workflow is configured to create records of the same type as the workflow (e.g., creating deal records within a deal-based workflow), if the newly created record meets the workflow's enrollment triggers, you could create a loop in which newly created records keep creating new records.

Pattern clue: You see a growing sequence of nearly identical deals created over a short time window. While record source and create date properties can hint that records are created by automation, you can then check for an explicit infinite-loop warning in the workflow history to confirm.

You need to make your enrollment triggers and re-enrollment rules strict enough that the same record isn’t repeatedly re-created.

Workflow Not Checking for Existing Records

A workflow is set up to create a new record based on a trigger (like a deal being created or a company being added), but it doesn't include logic to check whether a record with that info already exists.

Pattern clue: You see multiple records for the same entity. The record's creation date with a workflow Record Source is later than that of the other duplicate. 

For example, A prospect submits a demo request form, creating a contact record. The next day, a sales rep creates a deal and associates it with the wrong company. A workflow designed to ensure that all deals have an associated contact triggered the creation of a new contact record using information from the deal, unaware that the contact already existed.

This indicates the workflow doesn't include branches to verify whether a record already exists before creating a new one.

Remediation actions needed:

1. Review the workflow configuration
2. Replace "Create record" actions with "Set property value" actions
3. Or, use unique identifiers via personalization tokens when using "Create record"
4. Add conditional checks: "If record with [name] exists, update; otherwise create"
5. Turn off re-enrollment if it's not needed
6. Test the updated workflow with sample data
7. Merge the existing duplicates after confirming the workflow fix

In Insycle, navigate to Data Management > Merge Duplicates, and select the record type.

In Step 1, set up your basic matching criteria (name, domain, email, phone number, etc.). Don't filter by Original Source here yet.

Click Find.

In Step 2, customize the column layout.

Add Original Source, Original Source Data 1, Original Source Data 2 (or the equivalents for your selected record type), and position them after the basic identifying fields.

Save the template with a name like "Standard Duplicate Check with Source Analysis."

Now review the duplicate groups under Step 2 and look for a pattern. For example, multiple duplicate groups show:

• Original Source = "Offline Sources"
• Original Source Data 1 = "INTEGRATION"
• Original Source Data 2 = "581872"

You also notice a timeline: Duplicates started appearing 2 weeks ago.

Use Preview mode under Step 5 to export a CSV for detailed review.

To determine what the integration ID represents, click a deeplink for one of the duplicate records. Then, in HubSpot, review the Original Source properties.

For this example, you learn that integration ID 581872 correlates to your Shopify sync.

There is a timing correlation with when the Shopify app was connected to HubSpot two weeks ago. This suggests that the new integration is creating new records rather than updating existing ones.

Navigate to HubSpot Settings > Integrations > Connected Apps > Shopify and review the sync configuration for the object type.

In the sync settings, you discover there is currently no matching key configured. 

To fix this, enable the email address as the matching field.

To test, add a contact in Shopify that already exists in HubSpot and verify that no new duplicates are created.

Return to Insycle's Merge Duplicates module.

Configure master record selection and data retention rules under Step 4.

Use Preview mode in Step 5 to test the merge configuration, then execute the final merge in Step 5 using Update mode.

Wait 1 week, then run the analysis template with the Original Source fields.

In Step 1, click the Filter button and add a filter for the "Original Source Data 2" field value of "581872."

Verify there are no new duplicates with Original Source Data 2 = "581872."

Set up automated monthly monitoring templates to catch any new duplicate patterns.

Original Source properties were introduced by HubSpot in February 2024 and backfilled for the previous four years. Records created before this period may not have Original Source values populated. Additionally, some older records within the 4-year window may have incomplete data. 

HubSpot's Record Source properties are available for all object types, including custom objects, so you can use this same approach for any object type syncing with Insycle.

You don't necessarily need to use them as matching criteria in Step 1, but it's highly recommended to include them in your Step 2 layout for analysis purposes. This provides valuable insights into duplicate patterns, even if you're not filtering by source.

Check the Original Source Data 1 and Drill-Down 2 fields. These often contain the integration name, integration ID, or other identifying information. In HubSpot's integration settings, you can correlate these IDs to specific connected apps.