[US only] Beta test the new bulk image upload feature in Square Catalog 📸

[KCfromProduct]

Hi sellers!

We’re thrilled to invite you to test out a new and exciting beta feature. This update will make it easier for you to bulk upload and automatically associate images to existing items in your Square Catalog via Bulk Edit. 

Interested in joining?

👇 Reply directly to this post below and we’ll get you set up so you can start testing over the next few weeks. Once you drop a comment below, we will acknowledge your opt in to this beta by liking your reply and will follow up with you individually with any additional information.

We look forward to hearing from you!

[JTPets]

Does it have a filter to show only items missing images?

[KCfromProduct]

Hi John, this feature does not include a filter for missing images. To better understand your workflow, do you typically use your item library first to identify which items don't have images, take / source the images, and then return to your item library to upload?

[JTPets]

My work flow no longer requires this filter as I have worked with AI to build me a catalog export tool to google sheets that can automate all of this for me (released publically on this forum and my github). However I have seen this feature asked for and needed this filter in the past myself.

I have 2500 items, and previous to my automation work, I myself have stuggled with square filters to get the level of details I want. Sorting through items to find the ones missing images is a tedious process with a large catalog. There are users who need that capability still, and are and have asked for it on the forums.

[dokidoki]

Chiming in but yes, absolutely for folks who have thousands of items like myself. 

[JoeTend]

A missing image filter is something I’ve been requesting as well. Filtering to missing images at the SKU/ITEM VARIATION level would be even better!

[dokidoki]

I would love to join! In fact, I have thousands of items and this is one of my biggest pain points that I do on a regular basis when getting new items. 

[JoeTend]

I’d love to join!

[TCSlaguna]

I would like to join. Thanks! 

[RavensTownBalti]

Yes,Thank you

[akramer]

Yes please.

[KCfromProduct]

Hi folks! Thank you for your interest in our beta. We're still putting the finishing touches on this feature, but should be ready to kick off next week. I will circle back once the feature is ready to be tested!

[barefootwood]

I'd like to participate as well please.

[KCfromProduct]

Good morning! The automated bulk image upload feature should now be available to you in Square Dashboard. To access to go:

1. Items > select the items you wish to upload images for > click the "Edit items" button in the bottom right
2. Click the "..." button at the top right and select "Bulk upload images"
3. Select all of the images you wish to associate to the items you selected

All feedback is welcome, but in particular we're interested in getting your thoughts on:

• What types of items you tried to upload images for (e.g. physical items, prepared F&B items, a mix, etc.):
• Approximately how many item you tried to upload images for:
• How many image mappings were incorrect:
• Why or why not you found this feature useful:
• What improvements you'd like to see to make this feature more useful:

A few things to note:

• An image is only associated with an item if we deem it a high confidence match; if an image is not mapped, it's because it did not pass our confidence threshold
• This feature does not currently support mapping multiple images to a single item or mapping images to variations (this is something we may explore in the future!)
• If an image is incorrectly mapped, you can update it by clicking the image tile > selecting "Library" in the Image library modal > sorting by date uploaded to find the correct image

Thank you for your time and please reach out with any questions!

[dokidoki]

I've been testing this extensively and its such a wonderful new feature. Here are my recommendations and things I've noticed.

1. Testing books that are the same name but different volume really confuses it. I'd say it works only 50% of the time. For example, if the books that are named: Sailor Moon Vol. 1 and Sailor Moon Vol. 2, it often will give Sailor Moon Vol. 1 the Vol. 2 image and vice versa. So it will still attribute an image, but not the correct Volume image.

2. Fixing the wrong image would be much faster if the default selection was "Sorting by Date Uploaded" rather than Alphabetical Order. Having to switch it to "Sorting by Date Uploaded" every single time is a pain and I don't see much value in Alphabetical Order.

3. I highly recommend that the logic include looking at the filename and mapping it that way as well, especially by GTIN. All my images are named after the GTIN, so it would always be a 100% correct mapping if it would map the file name of the image to the item GTIN. So if there was a way to also upload bulk images by image filename mapping that would be amazing. Maybe even as a separate method so people could choose the current AI method or image filename mapping.

4. This feature breaks every time I bulk upload some of the items, and then upload more on the same screen. It needs to be done all at once or it just will randomly start applying multiple images to a single item.

5. I've tested bulk uploading only physical items so far and specifically just books. It works really well as long as its not a similar name as mentioned above. I've tested doing a small number at once (4) and doing about 25 at once. I will continue testing with a larger amount and try items that are not books.

[kc2429]

Thank you so much for the detailed feedback!

One follow up question for you re 4 – does this happen every time you select some images, map, and then select more or just occasionally? I just tested on my end and wasn't able to reproduce. Do you happen to remember what items / images you were using when this happened? If so, could you let me know which ones they were so I can use them on my test account? No worries if you don't remember the exact combination. 

Also, just fyi:

1. We'll consider adding in the ability for our model to read any text that's included in the image in a future iteration
2. We'll be updating this in the coming weeks!
3. We're ways to incorporate filename into the mapping!

[dokidoki]

I'll test it early next week and get back to you because I'll be adding new inventory then! Thank you for the follow up. Looking forward to seeing this feature grow as its massively helpful and time-saving.

[Twiggy]

Can I have this please when it’s ready for UK