Is Your Shopify CSV Usable? Check SKUs, Variants, and Images
Use a practical acceptance check for catalog exports: preserve identifiers, reconcile variants, verify image associations, and keep currency evidence.
The export opens in Excel and the rows look tidy. Then you discover that 00042 became 42, a long product ID was rounded, and a product with several sizes has only one row.
Those errors affect what you can do with the file. Before using a Shopify catalog export, run an acceptance check that goes beyond whether the download button worked.
I built Catalog Prep around preserving observed values and exposing capture context. Use the checks below for its exports or any alternative you are evaluating.
1. Preserve identifiers as text
Keep product IDs, variant IDs, SKUs, and barcodes as text. Spreadsheet type inference can remove leading zeros, change date-like strings, or round long numeric values.
Catalog Prep's ZIP includes a text-typed catalog.xlsx for inspection. Its exact CSV preserves source text. If importing that CSV manually, choose column types explicitly and keep an untouched original.
Literal text can also begin with characters a spreadsheet interprets as a formula. Use the text-typed workbook or a text-configured import when inspecting source values. There is no reason to execute catalog text as a spreadsheet formula.
2. Reconcile products and variants separately
One product can have several sizes, colors, or other options. In Catalog Prep's CSV, each returned variant has a row with its parent-product context.
Consider this illustrative sample:
| Product | Variant | Source SKU |
|---|---|---|
| Jacket A | Small | 00042-S |
| Jacket A | Medium | 00042-M |
| Jacket A | Large | Missing |
This is one product and three variants. The missing SKU does not justify dropping the third variant. Use stable product and variant identities where available; a SKU can be absent or reused.
Compare several products with the public source and inspect variant-coverage warnings. The exported rows represent returned source data, not proof of every private merchant option.
3. Check gallery and variant-image relationships
A product gallery and a variant-specific image answer different questions. The gallery shows captured product assets; a variant association connects an asset to a particular option where the source provides it.
Inspect images.csv, the product gallery fields, and normalized JSON together. Confirm that links point to the intended product's images, and that a default image has not been silently presented as an observed variant image.
Image links are references to source-hosted files. Catalog Prep exports those links and available metadata; it does not download image binaries. Retain capture dates and confirm permitted reuse before moving assets into another workflow.
4. Preserve tags without inventing classification
Tags can support category research and supplier mapping when they are exposed. An empty public tag list should remain empty. Product titles and collection membership can inform your own analysis, but should not be relabeled as source tags.
Keep any category labels you create in separate columns. This lets a colleague distinguish what the store published from what your team inferred.
5. Attach currency evidence to prices
An amount without currency context is incomplete for cross-store comparison. Preserve the source price string and inspect currency status and page-currency hints.
Catalog Prep detects store currency from public metadata when available and keeps the page hint separately. It does not convert prices. If metadata is unavailable, missing currency stays unknown.
Likewise, availability is a public flag, not actual inventory quantity. Keep it separate from stock or sales calculations.
6. Read coverage before counting missing items
Inspect the selected scope, source URL, product and variant counts, limits, capture time, and status in the manifest. A partial capture is useful when its boundary is visible.
Comparing a collection export with a whole-store screenshot can create an apparent mismatch even when both accurately describe different scopes. The collection guide shows how to select the intended set.
Use the downloadable catalog acceptance checklist to record each check, the evidence, and the follow-up. It contains review prompts, not fabricated product data or automated pass results.
Keep source review separate from destination import
A usable observation file still needs mapping for the system that will consume it. Native Shopify import columns, destination identities, pricing rules, and asset permissions need their own review. Catalog Prep's export does not write products into Shopify.
For a first evaluation, install Catalog Prep, capture a small public collection, and check a few products across CSV, Excel, and JSON. Use the getting-started guide if you need the capture steps.