How to bulk list on eBay

CSV upload. Fast if you already have structured data — for example, inventory exported from another system. Useless as a time saver if creating the CSV means typing the same fields into a spreadsheet instead of a form.

Template duplication. Excellent for near-identical inventory: one hundred copies of the same book, or the same shirt in ten sizes. Poor for anything where category or specifics differ per item, since you end up editing every duplicated row.

Batch photo processing. Each item is identified from its own photo, mapped to its own category, and given its own item specifics. This is the only approach that survives mixed inventory, which is what most estate-sale, thrift, and collectibles sellers actually have.

Most serious sellers use more than one. Templates for the repeat items, photo processing for everything else.

Bulk listing goes wrong in predictable ways: the wrong category duplicated across a hundred items, missing required specifics rejected at publish, condition copied from a template that didn't apply, and titles that are identical except for one word.

Two habits prevent most of it. First, validate before publishing rather than after — check category validity, required aspects, and condition validity while the item is still a draft. Second, spot-check a random sample from each batch rather than trusting the whole run because the first three looked fine.

HumCity runs the validation automatically and surfaces failures as warnings in the editor. The spot-check is still on you.

In any batch there will be items the AI can't identify, categories it can't confidently resolve, or items missing a required specific that isn't visible in the photo.

The correct handling is to leave those as drafts and publish the rest. A batch workflow that blocks on the worst item is a batch workflow you'll abandon. Come back to the stragglers with a better photo or manual entry.

Think in monthly throughput. HumCity's free tier covers 50 listings or revisions per month, which suits a casual seller. Paid tiers cover 500, 5,000, and 10,000 per month for $10, $25, and $50. Revisions count the same as new listings, which matters if you reprice frequently — a catalogue-wide price sweep consumes allowance.

A little organization before the upload saves far more time than it costs. Group your photos by rough category before processing — cards together, electronics together, apparel together — even though the tool identifies each item independently. It makes the resulting review queue easier to scan, because similar items tend to need similar kinds of correction, and you can work through a whole group of the same type of fix in a row.

Naming or numbering your photo files sequentially before upload also helps you cross-reference a physical pile of inventory back to its draft if you need to double-check something mid-review, especially in batches of 100 or more.

At volume, the easiest way to introduce inconsistency is to attach different shipping or return policies to different drafts without noticing. Set a single default business policy set for each rough category of inventory — lightweight collectibles versus heavier electronics, for example — and confirm the batch is using the right one before you approve the run. HumCity applies your chosen business policies at publish time, so a five-second check per batch prevents a mismatched shipping profile from reaching dozens of live listings at once.

Batch processing shines when your inventory is mixed rather than uniform. If every item is the same SKU in a different size, a template or CSV import is faster and simpler — the per-item variation is small enough that a form-based duplication tool handles it well. The batch photo approach earns its keep when each item is a different product entirely: a card from one set, a phone from one manufacturer, a jacket from one brand, a LEGO set from one theme.

That's the inventory profile most resellers actually have after a sourcing run. Estate sales, thrift hauls, storage auctions, and card show pickups all produce mixed lots. The AI listing pipeline was designed for exactly that case, because it derives each item's listing from its own image rather than from a shared template that assumes similarity.

For sellers with highly uniform inventory — say, a shoe store restocking the same models — batch photo processing still works, but the gain over templates is smaller. Match the tool to the inventory shape.

The review queue at volume is a different skill than reviewing one listing at a time. Sort by category so you're checking the same kind of item in a run — your brain stays calibrated for the same field names and condition vocabulary, which is noticeably faster than switching categories on every item. Filter for flagged fields first and clear those, because flagged items are the ones most likely to need a manual fix before they can publish.

Publish in runs of similar items rather than one at a time. If you've reviewed and approved twenty cards from the same set, publishing them together is one action rather than twenty, and it's easier to confirm they all went live correctly than to track individual publishes across a long session.