Product import: release 2 need specification

Now that release one of product import is almost done, and people start testing and giving feedback, here is the place to collect feedback and wishlist for iteration 2 of the release plan :slight_smile:

The first level of need should be met, but it seems we can’t really considered the need as met before add up the image bulk upload, so it seems we will need a second iteration to consider the feature as done. But we still need to prioritize iteration 2 in the global priority list, so opening that wishlist and will then open an icebox.

New level of need

  • Hub managers need to be able to easily and quickly import a big list of products. Most of them were using other software before so will export those data with full description and attribute AND IMAGES. Today the product import feature doesn’t enable uploading images, so hub managers need to add them one by one, which… when you have 200 products is… annoying. So hub manager needs to be able to bulk import images as well.
  • We anticipate some issues with the “conflicts on non-updatable fields for existing products”. If you upload a variant that modify a field “shared” at the product level, it’s an error, but not super UX friendly to fix, you need to either go in your product list and manually correct or do it in the csv file before reimporting. A suggestion for improvement would be to open a modal if this error happens for “merge conflict resolution” asking for each case which value to keep that would replace the other value". But let’s wait for user feedback before anticipating too much the need for it…
  • I’m wondering how much of a bottleneck it is to have to create producers before doing the import. I think it’s satisfying as it is as you need to take time to write up the producer description, etc. And the whole profile creation modal enable you to do that quite smoothly. But if the need emerge, we might have to think on how to enable product import for producers who have no profile yet. Do you think that’s something we would need from the request you got from users?
  • Also there might be some other bottleneck that users will encounter while using the feature, that we will gather here.

Links to more details

Possible solution

Solution discussion for image upload:

  • “I think they would have to supply a link to the image and we would save it as an attachment. It shouldn’t be too hard to tack on” (from Matt)
  • “would they have to have already uploaded the file somewhere to create links? seems like having a file name and a target directroy on computer would be easier for many people?”(from Kirsten)

All is summed up in this story map: https://app.cardboardit.com/maps/guests/277acf1ab2333d9f83c3f358127f1c0f3be72537df3d4235c134c800d72086f5

@Kirsten @lin_d_hop @SineadOFNUK I know for UX and Germany this feature has been quite important. I’m wondering now that release 1 is almost done if we need to move forward and plan a release 2, and if yes, what are the remaining needs not yet covered by release one?

As it stands now (and at least my understanding of it) I don’ t think any user in Canada (except maybe the tech savy @CLFC) would be using release 1 of the product import. So I suggest we move onto release 2 quickly. And - @MyriamBoure - you peaked my interest when you wonder if maybe we can upload products for producers who don’t have profiles set up. Might the product uploader create a producer ‘shell’ - that might be filled in with a full profile later? This would be a great benefit to larger hubs. And I’m finding in many cases, producers only want to sell through the hub, they don’t want to sell direct through OFN anyway.

While a product importer would probably save us a TON of time in some ways, in other ways, I believe we’d rather go old school and create all the products from scratch. Between there being a lot of old/unused product listings and having options for things like variants vs. a ton of new products that SHOULD be variants, it would just be better overall for us to do it manually. So I don’t see release 1 vs release 2 being an issue here.

Is the ability to download a pre-populated csv with all the current product details in it already on the wishlist? It seems like doing an update would be difficult/slow if you can’t download the product info as it is? This isn’t actual user feedback, just my thought.

1 Like

Moved to https://github.com/openfoodfoundation/wishlist/issues/127