Why Shopify CSV imports corrupt products — and how to avoid it

The Shopify product CSV isn't really a spreadsheet — it's a set of instructions the importer applies directly to your live catalog. When an instruction is ambiguous, the importer doesn't pause to ask; it picks an interpretation, runs it across every matching row, and offers no built-in undo afterward. Most "Shopify CSV import errors" trace back to one of the mechanics below.

1. The Handle column decides which product gets overwritten

Per Shopify's CSV documentation, the Handle is what uniquely identifies a product during import — not the title or SKU. With "Overwrite products with matching handles" selected, any row whose Handle matches an existing product overwrites that product's fields. That one column is the difference between "update this product" and "create a brand-new one."

Two failure modes follow from this:

• Wrong or changed Handle → the wrong product is edited. If you rename a handle or a tool re-slugs your titles, the row no longer points at the product you intended.
• Blank or non-matching Handle → duplicates. A row whose handle matches nothing is treated as a new product, so you end up with two near-identical listings instead of one updated one — a frequent complaint among merchants importing supplier feeds.

2. Missing Option columns delete your variants

This is the one that produces the dreaded "Shopify bulk editor deleted variants" moment — except it's usually the CSV importer, not the bulk editor. Shopify's docs are explicit: when you update a product that has variants, you must include the Option1 Name and Option1 Value columns. If those columns are missing, Shopify creates a single new default variant and deletes the existing variants.

It gets subtler. Shopify also notes that changing an Option value deletes the existing variant IDs and creates new ones — which can detach inventory levels and orphan anything keyed to the old IDs. A correctly structured multi-variant product looks like this: the first row carries the full product, and each following row repeats only the Handle plus the next Option value.

Drop the Option columns entirely, and Shopify reads "this product has no options" — collapsing three variants into one default.

3. Blank cells overwrite real data with nothing

There's a counter-intuitive distinction in Shopify's overwrite logic that trips up almost everyone. When you overwrite by matching handle:

• A non-required column that is present but blank overwrites the existing value with blank. Per Shopify's example, an existing Vendor of "John's Apparel" becomes empty if your CSV has a Vendor column left blank.
• A column that is omitted entirely from the file leaves the existing value untouched.

So "Shopify import deleted my descriptions" usually means a Body (HTML) column was present but empty. The cure: only include columns you intend to change — every column you keep is one you're authorizing to be rewritten, blanks included.

4. Encoding and locale mangle your characters

Shopify accepts only UTF-8 encoded CSV files. Open an export in a program that saves in a different encoding, and accented letters, currency symbols, and quotation marks can come back garbled. Curly "smart" quotes are a frequent culprit and can trigger an "illegal quoting" error that fails or distorts the import. Two limits worth knowing: a product CSV can't exceed 15 MB, and Google Sheets defaults to UTF-8 — part of why editing in Sheets sidesteps the re-encoding trap.

5. The real problem: no preview, no undo

Each mechanic above is survivable on its own. What makes them dangerous together is that, as of 2026, Shopify's admin gives you no dry-run preview of what an import will change and no native undo once it runs. Shopify's documentation states that an import can't be cancelled after it begins and that there's no history of past imports — which is why their guidance is to back up your data before you import. By the time you see the damage, the previous state is gone unless you saved it yourself.

A prevention checklist

Run through this before every import, especially on a live store:

1. 1

   Export a backup first

   Export the products you're about to touch and save the file untouched. That export is your only realistic rollback path if the import goes sideways.

2. 2

   Confirm every Handle matches

   Spot-check that handles in your file exactly match existing products. Mismatches create duplicates; renames overwrite the wrong product.

3. 3

   Keep Option columns intact

   If a product has variants, always include Option1 Name and Option1 Value (and Option2/3 where used). Don't change Option values unless you mean to regenerate variant IDs.

4. 4

   Only include columns you mean to change

   Remove columns you don't intend to edit rather than leaving them blank — blanks overwrite, omissions don't.

5. 5

   Save as UTF-8 and test small

   Confirm UTF-8 encoding, replace smart quotes with straight quotes, and import a handful of rows as a trial before running the full file.

How a validate-before-write design closes these gaps

A manual checklist helps, but it still leans on you to catch every blank cell and stray handle by eye. SheetBridge is a pre-launch Shopify app being built around the failure modes above, with safety as the core design principle. It's waitlist-only — no live customers yet — but here's how its intended design maps onto each problem on this page:

If you'd rather edit your catalog in a familiar spreadsheet without the CSV minefield, our guide to editing Shopify products in Google Sheets walks through a safer workflow, and how to undo a bulk edit in Shopify covers your options when a change has already gone live.