How to Fix HubSpot Import Errors After a Bad CSV Upload

It happens fast. You prepare a CSV file, import it into HubSpot, and within a few minutes realize something went wrong. Maybe a column was misaligned, overwriting company names with phone numbers. Maybe the import created 3,000 duplicates instead of updating existing records. Maybe you accidentally imported a test file with garbage data into production.

The sinking feeling is universal. But the damage is fixable — if you act methodically instead of panicking.

This guide covers the most common import mistakes in HubSpot, how to diagnose exactly what went wrong, and step-by-step processes for fixing each type of error.

Common import mistakes

Wrong column mapping

The most frequent import error. During the import wizard, you map CSV columns to HubSpot properties. If "Phone Number" in your CSV gets mapped to "Company Name" in HubSpot, every imported record now has a phone number where the company name should be. This is particularly damaging because it silently overwrites correct data with wrong data.

How to spot it: Records have obviously wrong data in one or more fields. A company named "555-0123" is a dead giveaway.

Creating duplicates instead of updating

If you don't include a unique identifier (HubSpot Record ID or Email for contacts) or don't select "Update existing records" during import, HubSpot creates new records for every row — even if those records already exist. You end up with duplicate contacts, companies, or deals.

How to spot it: Your record count jumped dramatically after the import. You see duplicate records when searching for specific contacts or companies.

Wrong object type

Importing contact data into the company object, or deal data into contacts. This happens when someone downloads a template for one object type, fills it with data for another, and imports without noticing.

How to spot it: The imported records exist in the wrong section of HubSpot. Contact data appears as company records, for example.

Encoding and formatting errors

CSV files with special characters (accented letters, non-Latin scripts, emoji) can get corrupted if the file encoding isn't UTF-8. Currency symbols, date formats, and number formatting also vary between locales and spreadsheet applications.

How to spot it: Records have garbled characters (like "Café" instead of "Cafe"), wrong date values (month and day swapped), or number fields with unexpected values.

Importing a test or draft file

Someone imports a file they were still working on — incomplete data, test rows, placeholder values like "John Test" or "[email protected]." These records are now live in your CRM alongside real data.

How to spot it: Records with obviously fake data ("Test Company," "Jane Doe 2," placeholder email addresses).

Overwriting good data with blanks

If your CSV has empty cells for certain properties and you mapped those columns during import, HubSpot may overwrite existing values with blanks. This depends on your import settings — HubSpot gives you the option to skip blank values, but if you didn't check that box, your data gets erased.

How to spot it: Records that previously had values in certain fields now show those fields as empty.

Step 1: Stop the bleeding

Before fixing anything, prevent further damage:

Don't import anything else until you've resolved the current issue. Additional imports on top of corrupted data make the problem harder to untangle.

Notify your team. Let sales, marketing, and service teams know that there was an import issue and data may be unreliable for certain records. This prevents reps from making decisions based on corrupted data.

Note the import timestamp. Go to Contacts > Import in HubSpot and find the import that caused the problem. Note the exact time it was processed and the number of records affected. This gives you a precise scope for the cleanup.

Step 2: Diagnose the damage

Review the import log

In HubSpot, navigate to the import that caused the problem. HubSpot provides:

• Total records processed: How many rows were in the file
• Records created: New records that were added
• Records updated: Existing records that were modified
• Errors: Rows that HubSpot couldn't process (and why)

The records created vs. updated breakdown tells you immediately whether duplicates were created (high creation count when you expected updates) or existing records were overwritten (high update count).

Check the column mapping

If you can access the import settings (sometimes visible in the import details), review which CSV columns were mapped to which HubSpot properties. This tells you exactly which properties were affected.

If you can't see the mapping, compare a few affected records against the original CSV file. Find a specific record in both places and check which CSV values ended up in which HubSpot properties.

Determine the scope

Create a list of affected records. If the import created new records, you can filter by "Create date is [import date]" and "Created by import" to isolate them. If the import updated existing records, you may need to use the property history feature to identify which records were modified.

For a quick scope check: compare your current record counts against what they were before the import (if you have that data). A sudden jump of 3,000 contacts on a day when you expected 0 new contacts is a clear indicator.

Step 3: Fix the damage

The fix depends on the type of error.

Fixing wrong column mapping

If properties were overwritten with wrong values, you need to restore the correct values. There are a few approaches:

If you have a backup export from before the import: Import the backup file and map it correctly. Use the Record ID column to match existing records and overwrite the corrupted values with the correct ones.

If the original CSV has the correct data: Re-import the same file with the correct column mapping. Again, use Record ID to match existing records. Make sure to check "skip blank values" if the original file has empty cells.

If the original data is in the wrong property: If "Phone Number" was mapped to "Company Name," the phone numbers are now in the Company Name field. You might be able to use a search-and-replace approach to identify records where the Company Name contains a phone number pattern (using regex) and clear or fix those values. Then re-import the original file with correct mappings.

Fixing duplicate records

For duplicates created by an import that should have updated existing records:

Delete the duplicates. If the import created entirely new records (rather than updating existing ones), you can select and delete the new records in bulk. Filter by create date and import source to isolate them.

In HubSpot, go to the relevant object type's index page, filter by "Create date is [import date]," select the records (you can use "Select all" for the filtered view), and delete them.

Merge the duplicates. If the import created records that now have some data that your original records don't, you may want to merge rather than delete. This preserves any new data from the import while eliminating the duplicate. Merging is more time-consuming since HubSpot only supports one-at-a-time merging.

Fixing wrong object type

If data was imported into the wrong object type, you'll need to:

1. Export the incorrectly created records
2. Delete them from the wrong object type
3. Re-import into the correct object type

There's no "move to different object" feature in HubSpot.

Fixing encoding errors

Re-open your original CSV in a text editor (not Excel, which can re-introduce encoding problems) and save it with UTF-8 encoding. Then re-import with the correct encoding. Most modern spreadsheet applications offer a "Save as UTF-8 CSV" option.

For records already corrupted, you'll need to identify the specific characters that were garbled and replace them. Here's what cleaning up artifacts looks like — in this case removing stray HTML tags:

A search-and-replace operation targeting common encoding artifacts (like "é" which should be "e" with an accent) can fix these in bulk.

Fixing blank overwrites

If good data was overwritten with blanks, restoration depends on whether you have the original data:

If you have a pre-import backup: Import the backup to restore the original values.

If you have property history: HubSpot stores property change history on each record. You can see what a property's value was before the import. However, there's no bulk way to restore previous values from the property history — it's a one-record-at-a-time operation.

If neither: The data may be lost. This is the most painful import error because blank overwrites are difficult to detect and impossible to fix without a backup. This is why we recommend always exporting a backup before any import.

Step 4: Verify the fix

After applying your fix, verify that the data is correct:

1. Spot-check 20-30 records across different segments. Open each record and verify that properties have the correct values.
2. Run key reports. Check that your most important dashboards and reports look correct. If a report showed 5,000 companies in the tech industry before the import and now shows 200, something is still wrong.
3. Test workflows. If the corrupted data triggered any workflows (send email, create task, change lifecycle stage), check for unintended actions and clean those up too.
4. Check integrations. If HubSpot syncs with other systems, verify that corrupted data didn't sync outward. You may need to do cleanup in connected systems as well.

Preventing future import errors

Always export a backup first

Before any import, export the records you expect to be affected. This gives you a restore point if things go wrong. The 5 minutes this takes has the potential to save hours of cleanup.

Use a test import first

If your CSV has 5,000 rows, extract 10 rows into a separate file and import that first. Verify that the mapping is correct and the data looks right. Then import the full file.

Always include Record ID

When updating existing records, include HubSpot's Record ID column in your CSV. This is the most reliable way to match imported rows to existing records. Email-based matching works for contacts but can fail for other object types.

Check "Skip blank values"

Unless you specifically want to clear property values, always check the "Don't overwrite existing values with blank values" option during import.

Double-check column mapping

During the import wizard, carefully review every column mapping. HubSpot auto-suggests mappings based on column header names, but these suggestions aren't always correct. A column named "Company" might get mapped to the wrong "Company" property.

Use consistent file formatting

Standardize your import process: always use UTF-8 encoding, always use the same date format, always use the same number formatting. Document these standards for your team.

A bad import is stressful, but it's recoverable. The key is to diagnose precisely what went wrong, apply the right fix for that specific type of error, and then put safeguards in place so it doesn't happen again.

---

Need to clean up after a bad import? SearchReplace lets you find and fix corrupted property values across all your HubSpot records, with preview and one-click rollback.