Checking bulk catalogue operations for growing stores

Test a bulk catalogue route by tracing one representative change from source through job results to the customer-facing product. Storing a large range does not prove staff can safely update prices, availability and details before a trading deadline.

Define the job and expected result

Choose a real task: a supplier price update, seasonal range release or stock correction. Record each field's source, the products and variants involved, the completion window and who approves the result. Include one item that must stay unpublished and one that must not change.

Before the full job, prepare a small sample: a new product, an existing product, a variant, a deliberately invalid row and a product whose stock comes from another system. State the expected price, visibility and saleable item identity for each. Use the installed version and actual field mapping.

Match the update route to the work

Shopify offers asynchronous GraphQL bulk mutations for bulk product imports.

In API versions 2026-01 and later, each app can run up to five bulk mutation operations per shop at once. A JSONL input file cannot exceed 100 MB, and an operation running beyond 24 hours fails. An accepted operation still needs its result file inspected.

WooCommerce has a built-in product CSV importer for creates and updates. An update normally matches an existing product by ID or SKU.

Large imports depend on server memory, upload limits and processing time.

Adobe Commerce supports asynchronous REST write endpoints that queue accepted requests and return an operation identifier. A consumer must process the queued messages.

The endpoint route differs between Adobe Commerce Cloud or on-premises projects and Adobe Commerce as a Cloud Service. Queue acceptance does not establish that each product update completed, so the implementer needs to show the completion and error record for the proposed deployment.

Bulk Import Capabilities Across Platforms (Australia)

Inspect the result and exceptions

Keep the input version, mapping, operator, start and finish times, and the import or job identifier. Compare intended creates, updates and skips with reported results, reviewing creates separately from updates. Review errors and sample completed records across the range, especially variants and unpublished items.

Then inspect the storefront and a sample basket or order path. Does the intended price appear? Does an unpublished product remain unavailable?

Check whether a connected stock system has overwritten a manual change, or vice versa. If another system owns stock, coordinate the change at that source.

Retain the first result after a correction. Approve the route when staff can identify what changed, what failed and how to repair an individual product within the required time.

Pros and Cons of Bulk Operations in Australian E-commerce

Bulk Operation Performance Metrics (Australia)