How to Turn an Excel Product List into an Online Catalogue

Back to the journal / E-commerce

By KORBI · 27 October 2026 · 3 min read

A product spreadsheet is a useful starting point, not a finished catalogue. Clean the source data before mapping it to a searchable customer experience.

An Excel file often contains the first usable version of a company’s product catalogue. It can also contain duplicate rows, mixed naming, empty cells and several ways to express the same variant. Putting the spreadsheet online without checking it moves those problems to customers. Begin by making the source understandable and repeatable.

1. Keep a safe copy and identify the owner

Save a read-only copy of the original file. Ask who owns each column, where product changes come from and how often the list is updated. If two teams maintain separate sheets, agree which one is authoritative before merging them. A named owner and a change process prevent later imports from overwriting a correction with an older value.

2. Create stable identifiers and a field dictionary

Every record needs a product identifier that stays consistent when a name or category changes. Variants need their own relationship to a parent product and an identifier for the specific combination a buyer selects. Define each column in plain language: what it means, the accepted format, whether it is required and an example of a valid value. Commerce platforms likewise rely on defined CSV fields and specific product and variant rules; inspect the destination system’s current import template before mapping data.[1][2]

3. Clean, validate and map the rows

Check for duplicate identifiers, accidental line breaks, inconsistent decimal formats, misleading units and values placed in the wrong column. Normalize category names and attribute values, but keep the original value beside the cleaned one until an owner approves the mapping. Do not silently drop a column because the first import template has no matching field; decide whether the information belongs in an attribute, document, product description or a separate system.

  • Require product ID, customer-facing name and category.
  • Use consistent units and vocabulary within each product family.
  • Check that each variant points to an existing parent item.
  • Flag incomplete records for a person rather than inventing a value.
  • Keep a clear error report so corrections can be made in the source.

4. Prepare images and technical documents

Match each file to a stable product identifier, approve its use and set an image priority. Check that dimensions are adequate, filenames do not collide and images show the correct variant. Product sheets and certificates should have a clear owner and update date. Write alternative text that describes the useful information in the image; do not stuff product keywords into it.

5. Design the catalogue around buyer questions

Create categories and filters around the way a buyer narrows a choice. A product family might need size, finish, material or compatibility; another needs colour, model and fit. Avoid adding every spreadsheet column as a filter. Show the most useful product details first and move technical documents to a clear secondary place. Test a few real product searches with the sales team before importing the full list.

6. Run a small import, then reconcile it

Start with a representative sample: one ordinary product, one product with multiple variants, one with several images and one record that is intentionally incomplete. Compare each result on the site against the source file. Test search, filters, variant selection, quotation requests and any bulk-update behavior. Only expand the import after the mapping and error process are understood.

Illustrative architecture for 2,000 products

For a catalogue of 2,000 products, a practical starting design is one authoritative product table, a separate variant relationship where needed, a controlled set of categories, a file store for approved images and documents, and a quotation request that references product and variant IDs. A scheduled import can validate the source, publish accepted changes and report rejected rows for review. If the business later needs current stock or customer-specific prices, those should be connected to the system that actually owns that information. This is an example architecture, not a claim about a completed KORBI project.

Sources & further reading

  1. Importing products with a CSV file — Shopify Help Center
  2. Product data specification — Google Merchant Center Help

Continue reading