How can I import recommendations in bulk?

Recommendify supports uploading custom recommendations in bulk with a CSV file. You can link a parent product to multiple child products so that when a customer views the parent, the children are shown in the recommendations.

Before getting started, you might want to download any custom recommendations you have created by hand. To do that, log in to Recommendify through your Shopify Apps page, click the "Products" button and select "CSV Upload" from the dropdown.

 

 

That will take you to the CSV upload page. On this page, click the "download the existing recommendations" link. A CSV file will be emailed to the address associated with your store.





The email will look something like this:



Click the link in it to download the CSV file.

In this example, we can see that the product with the handle "milton-jacket-navy" is already linked to the products with the handles "casual-indigo-shirt", "david-shirt-blue" and "david-shirt-black". Likewise, "dress-1" is linked to "bell-dress" and "sleeveless-dress".

In general, the first column is the handle of the parent product and the subsequent columns are the handles of the child products. When a customer views the parent, the children will be shown.

A few things to note about the CSV file:

  • Each cell contains a single product handle.
  • Each row has at least two horizontal cells, but can have as many as you want.
  • If a product is linked to four other products, then there is a total of five horizontal cells in that row.
  • Blank cells are ignored.
  • Spaces before and after product handles must be removed.

Ok, so let's say we have another product with the handle "3-4-sleeve-kimono-dress-coral". When customers view that product, we want them to be recommended the "sleeveless-dress". We simply make another row in the CSV as follows:

Then we upload save that and upload it on the "CSV Upload" page.

 

 

Note that ticking "wipe existing custom recommendations" will delete all custom recommendations in the database before creating new ones from the CSV. This may be useful if you want to maintain one canonical version of your recommendations in a CSV file, or if you have uploaded incorrect recommendations previously. If you don't check this box, the default behaviour is additive - the existing recommendations will be left alone and any new ones in the CSV will be created.

After the upload has been completed, you will be taken to a page where you can see the details of the upload. There may be warnings associated with the upload, indicated by an icon in the warnings column. You can click on it to download a summary of what went wrong with the upload.