Following these best practices will help you avoid common pitfalls, speed up your imports, and prevent accidental data loss.
Step 1: Back Up Your Data First
Before starting any import that will update existing data, always create a backup by exporting your current data. This is your safety net if something goes wrong.
Why This Matters
Easy recovery: If the import doesn't go as planned, you can restore your original data
Reference point: Compare the before and after to verify changes
Audit trail: Keep a record of what your data looked like before the update
Peace of mind: Import with confidence knowing you can undo changes if needed
How to Back Up
Open Altera in your Shopify Admin
Start a new export for the data type you're about to import (Products, Orders, Customers, etc.)
Export all columns - Don't filter or limit fields; export everything
Download and save the file in a safe location with a clear filename like
products_backup_2025-10-28.xlsxVerify the export - Open the file and confirm it contains all your data
When to Skip This Step
You can skip the backup if you're:
Creating entirely new items (not updating existing ones)
Working on a development or test store
Importing data that can be easily recreated
For everything else, especially on production stores, always back up first.
Step 2: Test with a Small Batch
Before running a full import with hundreds or thousands of items, always test with 1-2 items first. This simple step can save you hours of troubleshooting and prevent data issues.
Why This Matters
Catch formatting errors early: Identify column mapping issues, data type problems, or validation errors before they affect your entire dataset
Verify the results: Confirm that your data appears correctly in Shopify exactly as you expect
Adjust your approach: Make changes to your spreadsheet or import settings before committing to the full import
Prevent bulk mistakes: Avoid having to undo or fix hundreds of incorrectly imported items
How to Test
Create a copy of your spreadsheet
Keep only the header row and 1-2 data rows
Upload and import the test file
Check the results in Shopify Admin
Review the import results file for any warnings or errors
Once verified, proceed with your full import
Remove Unnecessary Columns
When updating existing data, only include the columns you actually need to change. This is one of the most important best practices for safe and efficient imports.
Required Columns
Always include at least one identifying column so Altera can find the correct items to update:
ID (most reliable, preferred for updates)
Handle (for products, collections, articles, etc.)
Name (for orders, draft orders)
Email (for customers)
SKU (for product variants)
Columns to Update
Include only the specific data fields you want to change. For example:
Good - Updating product prices:
Handle | Variant Price blue-t-shirt | 29.99 red-hoodie | 49.99
Bad - Including unnecessary columns:
Handle | Title | Body HTML | Vendor | Variant Price blue-t-shirt | | | | 29.99 red-hoodie | | | | 49.99
Why This Matters
Removing unnecessary columns provides several important benefits:
1. Faster Imports
Smaller file sizes upload and process more quickly
Less data to validate and transfer to Shopify
Reduces API calls and processing time
2. Prevents Accidental Data Loss
Many columns have special behavior when left blank:
Blank metafield columns delete the metafield - If you include a
Metafield: custom.warrantycolumn but leave cells blank, those metafields will be deleted from your products (see Metafields documentation)Blank values may clear existing data - Some fields interpret blank values as instructions to clear the existing data
Empty image columns can remove images - Including image-related columns with blank values may unintentionally remove existing product images
3. Clearer Intent
Your import file clearly shows what you're changing
Easier to review and audit before importing
Reduces confusion about which fields should be updated
4. Better Error Tracking
Easier to identify which column caused an error
Simpler to fix and re-import failed rows
Less data to review in error messages
Example Scenarios
Updating product SEO titles:
Handle | SEO Title blue-t-shirt | Buy Blue Cotton T-Shirt - Free Shipping red-hoodie | Premium Red Hoodie - Organic Cotton
Updating customer tags:
Email | Tags [email protected] | VIP, Wholesale [email protected] | Retail, Newsletter
Updating order tracking numbers:
Name | Lineitem Fulfillment Status | Lineitem Tracking Number #1001 | fulfilled | 1Z999AA10123456784 #1002 | fulfilled | 1Z999AA10123456785
Understand Blank Value Behavior
Different column types handle blank values differently. Understanding this behavior helps prevent unintended deletions.
Columns That Delete When Blank
Metafield columns: Blank metafield values will delete the metafield from the resource
Tags: An empty Tags column will remove all tags
Some text fields: Certain descriptive fields may be cleared
Columns That Ignore Blank Values
Most standard columns (like Title, Vendor, Product Type) will skip updates when the cell is blank, leaving the existing value unchanged.
Best Practice
If you're unsure how a column handles blank values:
Test with 1-2 items first (see above)
Check the field reference documentation for that data type
When in doubt, simply remove the column from your spreadsheet
Use the Correct Command
The Command column controls how Altera handles each row. Choose the right command for your use case:
MERGE (recommended for most updates): Updates existing items or creates new ones if not found
UPDATE: Only updates existing items; fails if the item doesn't exist
NEW: Only creates new items; fails if the item already exists
REPLACE: ⚠️ Completely deletes and recreates the item with only the data in your file (use with extreme caution)
DELETE: Permanently removes the item from your store
IGNORE: Skips the row entirely
For most update scenarios, MERGE is the safest choice.
Verify Your Data Before Importing
Check for Common Issues
Before uploading your file:
Required fields are present: Products need Title, Orders need line items, etc.
Data formatting is correct: Prices are numbers, dates follow ISO format, boolean values are true/false
No hidden columns or rows: Hidden data will still be processed during import
Column headers match exactly: Use the proper field names from the field reference
Use the Preview Screen
After uploading but before starting the import:
Review the data type detection
Check the column mapping
Verify the row count matches your expectations
Look for any warnings or validation messages
Note the Analysis ID (shown with an "A_" prefix) - you can copy this ID to reference when contacting support about validation issues
Review Import Results
After every import:
Download the results file: Contains success/failure status for each row
Check the Import Comment column: Shows specific error messages or warnings
Verify in Shopify Admin: Spot-check a few items to confirm data appears correctly
Review failed rows: Fix any errors and re-import just the failed items
Keep Your Import Files
After running an import, save the original spreadsheet you uploaded. This helps you:
Track what changes were made
Re-import if needed
Reference the exact data that was imported
Debug any issues that arise later
Common Mistakes to Avoid
❌ Skipping the Backup
"I'll just be careful and won't need a backup" - Always export your data first, especially on production stores!
❌ Skipping the Test Import
"I'll just import all 5,000 products now and hope it works" - Always test with 1-2 items first!
❌ Including All Exported Columns for a Small Update
When you export products, you get dozens of columns. Don't include them all when updating just prices - remove the unnecessary ones.
❌ Using REPLACE Instead of MERGE
REPLACE deletes the entire item and recreates it. Use MERGE for updates unless you specifically need to wipe all existing data.
❌ Leaving Metafield Columns Blank Unintentionally
If you export products with metafields, then edit only the prices, remember to either remove the metafield columns or fill in the existing values. Blank metafield cells will delete those metafields.
❌ Not Checking the Results File
The import might complete successfully but have important warnings or partial failures. Always download and review the results.