Importing a CSV
Products › Import / Export. The file is a .csv up to the size the page states, and commas, semicolons and tabs are all read, so a spreadsheet exported anywhere in Europe works without being converted first.
- Press Catalogue file and choose the file. It is read in your browser and summarised: how many products, how many items and how many pictures it holds, and how many rows those came from.
- Look at the preview. It is the first rows as they will be imported, not as they are in the file.
- Check the column matching if anything looks wrong, and fix it. See below.
- Press Import n products. Rows are matched on SKU, so importing the same file twice updates rather than duplicates.
Products are grouped by their handle where the file has one: rows sharing a handle become one product with several variants, which is the shape Shopify and most shop exports use. A file with no handle column becomes one product per row.
Matching the columns
Change which columns are used opens the mapping. Each of your columns feeds one field, and each row shows an example value from your own file - which is the point: a column of costs matched to Price imports quietly and wrongly, and the example is what catches it.
- Columns matched by similarity are marked as such. They are worth a glance rather than trust.
- - do not import - leaves a column out entirely. Anything left out is listed above the preview so it is never silent.
- A file with two columns of the same name says so, because only one of them can be used.
Errors and warnings
The preview counts both. An error is a row that cannot be imported; a warning is one that will be imported with something adjusted. Each is listed with its line number and what happened.
After a run, Download the rows to fix gives you a spreadsheet holding only the rows that did not import, with the reason beside each one. Fix them there and upload that file: it updates what is already in rather than duplicating it.
If an import is interrupted, part of the file may already be in. Importing it again is safe for the same reason - rows are matched on SKU.
Every field you can import
| Field | Notes |
|---|---|
| Group (handle) | Rows sharing one become a single product. |
| Product title | Required for a new product. |
| Description | HTML or plain text. |
| Brand | The maker. |
| Type | The source shop's own word for a shelf. |
| Tags | Comma separated. |
| Status | Active, draft or archived. |
| SKU | The key everything matches on. |
| Option 1 to 5, name and value | The axes and this row's point on them. |
| Barcode, EAN, UPC, ASIN, part number | Stored in one common form so they scan alike. |
| Weight and weight unit | Read in the unit the file names. |
| Needs shipping | Off for downloads, services and gift cards. |
| Cost | What you pay. Feeds margin. |
| Selling price and compare-at price | In your reporting currency. |
| Stock on hand | Counted into your default warehouse. |
| Picture address, variant picture, order, description | Pictures by URL, in the order the file gives. |
Importing from a connected shop
Open the connection and press Import products on the Import and sync tab. The run is a job: it appears immediately, reports how many it has read as it goes, and survives you closing the page.
Some channels prepare a catalogue before they will hand it over, and the screen says so rather than looking stuck. What each import does with pictures is a per-connection setting, described on Connecting a channel.
What arrives: products, variants, options, prices, stock as the shop reports it, pictures in their order, and the shop's own word for a category, which the Categories screen turns into yours. Import the same shop again whenever you like; it updates on SKU.
Exporting your catalogue
Prepare the file writes a CSV of everything you hold, in the same shape this page reads back. It runs in the background, keeps going if you close the page, and reports progress on the Jobs screen. When it is ready, press Download.
There is no cap on how many products it covers. A large catalogue takes a few minutes, and a previous file stays downloadable until the new one finishes.