The Bulk operations tab lets you update many products at once instead of editing them one at a time. Under Select your operation, you will find four operations:
Catalog data. Update product fields (name, brand, GTIN, and more) from a CSV file.
Product images. Upload real product photos by item code.
AI images. Generate product images with AI for products that are missing them.
Match approval. Approve review-required matches in batches by setting a confidence threshold.
All three apply as your own edits. They take priority over matched Content Library content, affect only your catalog, and never change another distributor's data.
Catalog data: update fields from a CSV
Use this to change product fields across many items at once.
Choose Catalog data, then click Download Template for a CSV with the correct headers.
Fill it in. item_id is required, since it is how each row is matched to a product. item_name, brand, description, pack_size, gtin, manufacturer, and category are optional.
Upload the file (.csv only, up to 20 MB), then review the Confirm upload step.
When it finishes, you see Matched, Updated, and Skipped counts, and a report is saved to Reports. Use Download Errors for the list of item IDs that did not match.
Key rules:
Updates existing products only. Rows with an unknown item ID are skipped, and it does not create new products.
Only the cells you fill in change. A blank cell is left unchanged, so you cannot clear a field this way. To clear a field, edit the product directly.
Product images: upload real photos in bulk
Use this to add your own product photos across many items at once. Content Hub matches each image to a product by its item code, in one of two ways:
Name folders after item codes. For example, a folder named ABC-123 containing that product's images.
Name image files after item codes. For example, an image file named ABC-123.jpg.
Then drag and drop your folder or files to upload them.
Up to 8 images per item code.
Item codes are checked against your catalog, and unknown codes are skipped.
Uploading a large batch of images
There is no cap on how many images you can put through one session. Drop a folder of a few thousand photos and Content Hub works through them in stages rather than asking you to split the folder up yourself.
Note: Images over 2 MB are compressed before upload. Anything at or under 2 MB uploads exactly as it is.
If some files fail
The run pauses rather than pushing on, and the panel says Some files did not upload with the count that failed. Two choices:
Retry failed files processes them again with fresh upload links.
Skip failed files drops them and continues with the rest.
Everything uploaded up to that point is already saved either way.
Leaving or canceling mid run
Navigating away or closing the tab while an upload is running prompts you first: Leave this page? The warning explains that leaving cancels files that have not finished uploading, and that files already uploaded will still be saved. Choose Stay to go back to the run, or Leave to abandon it.
Canceling from inside the tool behaves the same way. Files that had already finished uploading are kept and the result is titled Upload canceled.
Reading the results
Each file resolves to one of these:
Uploaded. The photo is now on the item.
Already uploaded. That exact photo was already on the item, so nothing changed. Skipped. You chose to skip it after a failure, or it was set aside during checking. Failed. It did not upload. The reason is in the Details column. Canceled. The run ended before it uploaded.
Not found. The item code did not match anything in your catalog.
A file that matched several items reports per item, so one photo can be Uploaded on one item and Already uploaded on another. The summary counts items, not files.
AI images: generate images with AI
Use this to create images for products that do not have any. Choose AI images, then pick a type of operation:
Create new image. Generates a brand-new AI product image from the item's name and details.
Enhance image. Improves an existing product image, for example by cleaning up the background. Requires a product that already has an image.
Regenerate previous. Creates a new variation of a previously generated AI image. Available for products that already have an AI image
Then use Search products to add to pick the products, generate the images, and review each result to accept the ones you want.
Good to know:
Generation is driven by the item name and the details you provide, not by any other source. If the name is vague, add detail to get a better image.
AI images work best for generic, unbranded products. For branded items with a barcode, matching usually pulls the real product photo.
Match approval: approve review-required matches in bulk
Match approval clears the review queue in batches instead of one product at a time. You set a confidence threshold, and every suggested match at or above it is approved together.
Under Select your operation, choose Match approval.
Setting the threshold
The Confidence threshold slider opens at 90%. Everything at or above the threshold is approved; everything below it stays in review. The slider covers the full 0 to 100 scale in steps of 5, and the number field next to it takes any whole number. Your account's own flag threshold is shown under the slider as "Your flag threshold: {n}%", for reference.
As you move the threshold the count updates to "{n} items at or above {pct}%". If nothing qualifies you see No matches at or above {pct}% with "Lower the threshold to find suggested matches to approve."
Note: The threshold here does not change your account's Direct Match or Suggested Match settings. It only decides which flagged items this batch approves. Changing what goes to review in the first place is still done in Settings.
Reviewing before you approve
Click Review {n} matches to see what is included. The table lists Product, Confidence, Suggested Content Hub match, and Flag reason, 200 rows to a page. Untick any row to exclude it, and the running count shows how many are approving and how many are excluded.
Approving
Click Approve {n} matches. You land on Approved {n} matches with "They are now matched and no longer need review."
If someone else resolved an item while the batch ran, that item is skipped rather than overwritten, and the result says so: "{n} was skipped because someone else resolved it while this ran."
Click Approve more to go back to the threshold with the queue refreshed.
Note: Approving in bulk accepts the match Content Hub suggested, as it stands. It is the same outcome as accepting each one from the product's own page, so raise the threshold if you would rather eyeball the weaker suggestions individually.
If an approval does not finish
"Couldn't finish approving matches." with "Matches approved so far are saved. Try again to continue." Part of the batch went through. Retrying picks up the rest rather than redoing the lot.
"Too many bulk approvals in the last minute. Wait a moment, then try again." You have run several batches in quick succession. Wait, then retry.
How bulk changes behave
Whichever operation you use, the same principles apply:
Changes save as your own edits and take priority over matched Content Library content.
They apply only to your catalog, across all divisions, and never to other distributors.
Products are matched by your item ID or item code, and unmatched entries are skipped and reported.
Changes are not instant. Once a run finishes, updated content takes up to four hours to reflect in Pepper.
Frequently asked questions
Can bulk operations create brand-new products? No. All three operations update products already in your catalog. Unmatched IDs are skipped.
Will bulk changes overwrite my manual edits or other distributors' data? Bulk changes save as your own edits, so a filled value replaces the current value and takes priority over matched content. They apply only to your catalog and never affect other distributors.
Which should I use to get product images, Product images or AI images? Use Product images when you have real photos. Use AI images when you do not have a photo and want one generated. For branded, barcoded items, matching often already supplies the real image.
How long until my bulk changes show up in Pepper? Up to four hours after the run finishes. The Matched, Updated, and Skipped counts and the saved report tell you the run itself succeeded, so you do not need to wait on those to confirm it worked.