
You open a listing to lower the price, and eBay won't save it: the Brand field is empty. The brand is already in the title, twice in the description. eBay still wants the field, because the title and item specifics do different jobs. The title gets you into keyword search. Item specifics get you through the filters buyers click on the left.
| What they are | Structured fields: Brand, Color, Size, Model, MPN and more |
| Why they matter | A listing shows in filtered results only if the field holds a matching value |
| Fill first | Required fields, then recommended ones that are true for the item |
| Work by | One category at a time; the field list changes with the category |
eBay uses item specifics for keyword search, filters, category pages and external shopping search engines (eBay's item specifics requirements).
A quick test: a buyer searches "black running shoes", then filters Color = Black and Size = 10. Your listing "Black Running Shoes Size 10" with empty Color and Size fields drops out at the filter. The title worked for the search, not for the filter. For a buyer who filters, an empty Color field reads as "not this color".
Each field in a category has a label:
A clear photo of the label or model plate proves the value; the field makes it filterable. If your photos are small, you can preserve detail in enlarged product photos so the print stays readable on zoom.
| Label | If blank | Fill it when |
|---|---|---|
| Required | The listing may not publish | Always, with the real value |
| Required soon | Becomes required on eBay's date | Before that date, on listings you'll revise |
| Recommended | Misses filters and searches | The attribute is true for the item |
| Optional | Usually nothing blocks | A buyer would filter or compare on it |
| Does not apply | A warning can stay | The field really can't apply |
Required fields are a gate, recommended ones are visibility. Don't spend an hour on optional fields and then fail a revise on one empty required field. eBay's item specifics guidance also lets you dismiss recommendations that don't fit. eBay doesn't publish how much each field lifts sales, so decide the order by what you can see.
Order the work in Seller Hub:
Rule: a field earns its place when it's true, right for the category and something a buyer would filter on. Guessed values make a listing look complete and describe the wrong product. "Does not apply" is for fields that can't apply, like an ISBN on a handmade item, not for ones you haven't checked yet.
| Category | Usually required or key | Usually recommended |
|---|---|---|
| Phones | Condition, Brand, Model, Storage Capacity, Color | Network, RAM, Lock Status, Operating System |
| Apparel | Brand; Size and Color per variation | Material, Style, Department |
| Parts | Brand, MPN, Type, compatibility | Placement, material if printed |
| Books and media | Format, author, language | Publisher, genre |
Don't guess: a storage size or network the phone doesn't have, a size that isn't on the tag, fitment you haven't checked, an ISBN or UPC from another edition.
These are examples. The source of truth is the listing form for your exact category. Open one live listing and compare its form with the table: where they disagree, the form wins.
Identifiers (GTIN, UPC, ISBN, MPN) go in only when they match this exact item or variant. A code copied from another edition or storage size links your listing to the wrong product. Wording matters too: if eBay suggests "Blue" and you type "Navy", the filter may miss you. Pick eBay's suggested value. To see which values sold listings in your category use, check eBay sold listings. Shipping to Amazon warehouses as well? The same label discipline applies to Amazon FBA packaging requirements.
A shirt in red and blue, sizes S to L, with every color and size written in the description: buyers still can't filter for "blue, large", and there's no size selector. A description never feeds a filter. The values were in the wrong place.
eBay's developer guide to variations splits fields into two levels:
One listing with variations when the product is the same, only a variation attribute changes and the category supports variations: a hoodie in four sizes and two colors. Separate listings when items differ by model, compatibility or condition: two phone models don't share a listing because both are black, and new and used don't hide in one selector.
Supplier spreadsheets often use their own column names, like "Colourway" or "Edition". Map them to eBay's names before you upload.
| Method | Best for | Watch for |
|---|---|---|
| Seller Hub listing form | One or two listings | Fields differ by category |
| Seller Hub bulk edit | A filtered set in one category | A field from another category landing by mistake |
| File upload | Hundreds of listings | Rejected rows: test 10 first |
| API | Syncing from your own catalog system | Developer work |
Helium 10 and Jungle Scout are built for Amazon research; they can't tell you what eBay requires in a category this week. Never invent GTIN, UPC, ISBN or MPN values to clear a blank. Setting up from scratch? Line this up with how to set up an eBay Store.
Made-up phone "Harbor Note 64":
| Field | Before | After |
|---|---|---|
| Title | Harbor Note 64GB Unlocked Black | The same |
| Condition | Empty: won't publish | Used |
| Model | Empty | Note |
| Storage Capacity | Empty, only in the title | 64 GB |
| Color | Empty, only in the description | Black |
| Network | Empty | Unlocked |
| UPC | Copied from the 128 GB version | Removed |
Title and price didn't change, only the fields. Color and storage are now filterable, the listing publishes, and the wrong code is gone instead of replaced by a guess. This example shows the mechanics, not a sales result.
The five usual mistakes: the same values copied across categories, free text instead of eBay's values, "Does not apply" on a real attribute, variation details in the description, and uploading a full file without a test batch.
Complete specifics make a listing eligible for more filters. They don't guarantee sales and won't fix a wrong category, weak photos, a high price or low demand. Before editing, note impressions, clicks and conversion for those listings in Seller Hub; screenshot it on the day. After, compare the same listings over the same length of time. For a general approach, see how to analyze your ecommerce data.
SellerClaw connects to your eBay store and works with your real listings, inside Claude or ChatGPT or as an agent in the SellerClaw app and Telegram. It can find listings with empty item specifics, fill them from your store data and rewrite titles, and shows you the drafts for approval. You decide how much it may change on its own. Title keywords and paid promotion are separate jobs: see eBay Promoted Listings.
Example request: "Find my active listings in Cell Phones & Smartphones with empty required item specifics, fill them from the titles and photos, and show me the drafts before you publish."
Pick one category today, fix its required fields on ten listings and screenshot their traffic. Want help with the rest of the catalog? Start free with SellerClaw.