POS Catalog Import Guide 2026: Prevent Setup Errors

A successful POS catalog import is not simply a spreadsheet that uploads without an error. The resulting items must scan correctly, show the right variations, carry the intended prices and tax settings, and appear at the locations where you sell them. A technically accepted file can still create hours of checkout corrections if the mapping is wrong.

This 2026 buying guide compares Square item imports, Shopify product CSV imports, and assisted migration. It is for retailers choosing a new system or purchasing migration help, not a universal file template. POSadvice.com helps you compare POS systems; the providers supply their own software, hardware, and implementation services.

Research basis: official Square and Shopify documentation reviewed October 5, 2026. This is a documentation-based workflow comparison, not a hands-on import benchmark. Templates and supported fields can change, so download the current template from the account you will use.

Quick answer: test the difficult products first

Before buying migration services or signing a POS contract, ask for a small pilot import containing your hardest products. Include a leading-zero barcode, an item with multiple sizes and colors, a product sold at two locations, a tax exception approved by your adviser, and an item with an image. Those examples tell you more than a demonstration containing ten identical coffee mugs.

Define success in business terms: each barcode finds one intended variation, prices match the approved list, each location can sell the right products, and stock is accurate at the agreed cutover time. Require evidence for each check. A message that says “import complete” proves the processing finished, not that the store is ready to trade.

Catalog import options compared

Assisted migration is a service category, not a standardized feature bundle. A quote for “product migration” may not include stock by location, images, modifiers, customer balances, or historical sales. Compare written scope rather than assuming every seller uses the same definition.

Separate products, variations, barcodes, and stock

A product is the general item being sold. A variation describes a specific sellable choice, such as a navy shirt in medium. A SKU is your business identifier for a sellable item; a barcode is the value a scanner reads. The identifiers may overlap in a particular store, but they should not be treated as interchangeable without checking the destination system.

Inventory adds a different dimension: quantity at a particular location and time. Importing the shirt’s description does not prove that twelve navy mediums exist in the downtown shop. Ask whether quantities are included in the product file, imported separately, or loaded through another tool. Establish how sales during the migration window will be reflected.

Also distinguish a supplier’s code from your own SKU. If different suppliers use the same code for different products, a naive mapping can create collisions. Keep a field dictionary that states the source, destination, meaning, and expected example for every important column. This becomes the shared reference for your staff and the implementation provider.

What to verify with Square imports

Square’s bulk item import documentation describes spreadsheet imports for initial catalogs, changes to existing catalogs, and transfers between Square accounts. It recommends exporting and saving a distinct copy of an existing catalog before making changes. Preserve that original separately from the file you edit.

The same documentation explains that tokens identify item variations. When transferring items into a different Square account, it instructs sellers to clear token values while retaining the Token column and its header. That is a specific account-transfer instruction, not a general instruction to remove identifiers from every update. Ask your migration provider to distinguish an in-account edit from a cross-account transfer.

Square also documents location-specific column names and required import headings. Tax-column headings include the applicable percentage format, and multiple-location imports require the relevant enabled-location columns. Use the current template and confirm column matching before upload. A spreadsheet that looks readable to a person can still assign data incorrectly if its headings or mappings are changed.

Ask how import errors will be reviewed. Square describes both automatic updates and manual corrections when a file has errors. The person approving a correction should understand its effect on the catalog; speed alone is not a reason to accept every suggested transformation.

What to verify with Shopify imports

Shopify’s product CSV import guide explains an optional overwrite setting for products with matching handles. When overwrite is selected, values in included matching columns can replace existing product data. A blank optional field and an omitted column can therefore have different consequences.

For example, Shopify documents that a blank included Vendor field can clear an existing vendor value, while an optional column that is absent generally leaves its existing value unchanged. It also warns about dependencies between variation-related columns. Ask the provider to demonstrate a narrow update that changes only the intended field while preserving options, images, and other existing data.

Shopify warns that changing option values can replace variant IDs and break third-party dependencies. It also states that a started product CSV import cannot be canceled. Those details make a small rehearsal valuable. Confirm what “rollback” really means for your catalog and connected apps rather than assuming a universal undo button exists.

Product availability and location inventory deserve their own checks. Shopify’s guide discusses publishing new products to channels and using inventory CSV workflows for location inventory. Have the representative show that an imported item is both available in the intended selling channel and stocked at the correct location. A product visible in the admin is not enough.

Build a pilot file that exposes real mistakes

Use a deliberately varied sample, perhaps twenty to fifty representative products rather than the easiest first rows of your existing export. The appropriate size depends on catalog complexity. Include discontinued items, multiple variations, accented characters, long descriptions, and products whose images are hosted separately.

Preserve leading zeros in identifiers by treating those values as text in the spreadsheet. Check whether your editor converts long barcodes into scientific notation or changes codes into dates. Open the saved file again before upload and compare several identifier values against the original source. The visual appearance of a cell is not always proof of the saved value.

Include duplicate-looking names with genuinely different SKUs and a deliberate duplicate identifier that the provider should detect. Test whether warnings identify the affected row clearly and whether a partial import creates records before an error stops the remaining work. Record the process for correcting and rerunning a failed batch without producing duplicates.

After the pilot, scan physical labels, sell a test variation in an appropriate demonstration workflow, inspect the receipt, and verify a return lookup. Keep a checklist of expected outcomes beside the observed results. Where a live transaction is required, coordinate the provider’s authorized test process instead of improvising charges.

Keep purchasing and inventory workflows connected

Catalog decisions affect more than checkout. If purchasing uses supplier codes, costs, and case quantities, verify those records against the receiving process. Our POS purchase orders guide explains questions to ask about ordering and receiving; do not assume a clean product import automatically configures those workflows.

Bundles need special attention because the parent item and its components may be separate records with a relationship between them. Importing all of the names does not prove the component stock deductions work. See our POS bundle inventory guide for a separate demonstration checklist. Ask whether such relationships are imported, rebuilt manually, or handled by another application.

Leave customer credits, gift card balances, deposits, and historical transactions outside the product-import assumption unless they are explicitly included in the migration scope. They are different records with different reconciliation requirements. A supplier who can move product descriptions may still need a separate process for those balances.

Pros and cons of self-service importing

Pros

• Direct control: your team can inspect the mapping and understand how the new catalog is organized.
• Repeatable updates: a documented file process can help with future additions or price changes.
• Clear source records: saved exports provide evidence of the values used for the migration.
• Potentially lower service spend: a straightforward catalog may not need extensive specialist conversion.

Cons

• Staff time: cleaning identifiers, variations, and images can take longer than uploading the file.
• Large error impact: one misunderstood overwrite option can change many products.
• Incomplete scope: product, stock, channel, and integration records may need separate work.
• Recovery limitations: an exported file is useful evidence, but not a guarantee of complete restoration.

Assisted migration can be worthwhile when the provider demonstrates your difficult cases and accepts measurable completion criteria. It is less valuable when “migration included” means only uploading whatever spreadsheet you supply.

Compare migration quotes using measurable deliverables

Ask whether the quoted record count means products, variations, inventory rows, or all of them. Consider a hypothetical store with 300 styles, four sizes, and three colors per style. If every combination exists, that is 3,600 sellable variations, not merely 300 import records. Real catalogs may have fewer combinations, so count the actual source data before accepting a price.

Request separate prices for data cleaning, field mapping, image handling, pilot testing, full import, location stock, training, and corrections after launch. Specify who fixes source errors and whether a second import is included. Ask whether temporary duplicate products, app reconnections, or barcode reprinting would create extra work.

Put acceptance conditions in the proposal. For example: every pilot barcode resolves to the approved variation; all pilot prices match; location quantities reconcile to the signed cutover file; and known exceptions have named owners. Expand the final checks to all critical records and a risk-based sample of the rest. Do not replace those conditions with an unsupported promise of perfect accuracy.

Plan the final cutover and the first trading day

Choose a source-of-truth export time and define how later sales, receipts, and adjustments will be captured. Avoid editing the same master catalog in two systems without a reconciliation plan. Keep the original export, transformed file, mapping notes, exception list, and final approved totals together.

On the first trading day, assign someone to monitor scan failures, missing options, incorrect prices, and unavailable items. Log the specific SKU and correction rather than creating an improvised duplicate at checkout. Review repeated errors before changing a whole category or rerunning an entire import.

A good purchasing decision leaves you with a usable catalog and a process your team understands. POSadvice.com helps you compare POS systems and implementation quotes. Tell providers your product and variation counts, locations, current platform, required integrations, and hardest catalog examples.

Frequently asked questions

Does importing products also import inventory at every location?

Not necessarily. Product details, channel availability, and location inventory can use different fields or workflows. Ask the provider to demonstrate each requirement and reconcile quantities to the agreed cutover record.

Should I remove Square tokens before every catalog update?

No. Square documents clearing token values when transferring items to a different Square account while retaining the Token column. Do not apply that account-transfer instruction indiscriminately to updates within an existing account.

Can a Shopify CSV import overwrite existing product data?

Yes. With overwrite enabled for matching handles, included columns can replace existing values. Blank fields, omitted columns, and variation dependencies have different effects, so test the intended change with a small sample first.

What should an assisted catalog migration quote include?

Specify products and variations, field mapping, images, location stock, pilot testing, corrections, and final acceptance checks. Identify any separate work for bundles, integrations, customer balances, or historical transactions.