How fast you launch on a marketplace depends less on the number of products than on how cleanly the data is gathered. A listing missing a mandatory field gets sent back by moderation, and one a shopper can't filter for simply isn't seen. Here is how to prepare ahead of time.
Gather complete product information
A complete product record isn't a name and a price — it's the set of fields the listing name, specifications and search filters are built from. Each item needs:
- Brand and model — what shoppers search for and use to tell items apart;
- Category and type — determines which fields the platform makes mandatory;
- Material — a separate field for most clothing, footwear, furniture and homeware;
- Dimensions and weight — physical parameters and, where relevant, the size chart;
- Colour — picked from the values the category accepts, not free text;
- Package contents — what's in the box, especially when accessories sell separately;
- Key features — what sets this model apart from the next line in the price list;
- Supplier SKU or barcode — needed for stock reconciliation.
The gap between complete and incomplete shows on one price-list row.
| Field | Before (as supplied) | After (upload-ready) | |---|---|---| | Name | black tshirt XXL cotton | Men's cotton T-shirt, black, XXL | | Size | XXL | XXL | | Colour | blk. | Black | | Material | — | 100% cotton | | Category attributes | — | Gender: men's; Season: all-season |
The difference isn't tone, it's structure: the name follows the category template, the colour is picked from the platform's list, and the fields moderation requires are filled in.
Bring the data to one format
A single format for names, units and specifications across the catalog saves hours of fixing later. The same inconsistencies come up again and again:
- Units written three ways: "kg" in one row, "kilograms" in another, a bare number in a third. A weight or volume filter only works when every value shares one format.
- Size as a letter, a word or a number: "M", "Size M", "medium", "42-44" — the filter treats these as different values, even when they mean the same size.
- Colour as free text instead of a fixed list: "navy", "dark blue", "midnight blue" — one colour written three ways, none matching the category's filter.
- Mandatory attributes left blank: a size chart for clothing, voltage for electronics, minimum age for children's items. Without them a listing either doesn't publish, or sits lower in the list.
Each inconsistency looks minor alone; across a catalog of a few hundred items, they add up to a line-by-line cleanup.
Prepare quality photos
Sharp images on a light background from several angles are a baseline requirement almost everywhere. Photos are a common rejection reason:
- Supplier watermarks or logos on the photo — the platform reads them as a competing brand on the listing.
- Stray objects in frame: a hanger, another seller's price tag, the phone's own reflection.
- Resolution too low or the wrong aspect ratio — fine in a chat thread, blurry once stretched into the layout.
- A collage with text overlays instead of a plain photo — most platforms only accept individual images, no text on top.
- The photo doesn't match the variant — the same picture used for both the black and blue version, though colour is what the shopper filtered by.
In practice, "several angles" means: front, back or side, a close-up of a detail or label, and, when size matters to the choice, a shot with a scale reference or on a model.
Pre-launch checklist
Before uploading the finished file, check each point:
- Every item has brand, model, material, dimensions, colour and package contents filled in, or the field is marked not applicable.
- Units of measure are consistent across the catalog: one system for weight, one for volume, one for dimensions.
- Colour is picked from the category's list of values, not typed as free text.
- Mandatory category attributes are filled in for every item, not just part of the catalog.
- Photos cover several angles, have no stray objects or watermarks, and match the specific variant.
- Product names follow one template across the category, not copied from the supplier's price list as is.
When the supplier's file is the only source, and it's poor
Often there's no separate source data, only a supplier file with everything tangled together: names in the original language, size and colour packed into one cell, half the fields empty. The order of work is the same, just starting a step earlier: work out what's actually in the file before generating listings from it.
Split mixed fields into separate columns: size and colour in one cell are two attributes, not one string. Mark whatever is missing, and don't fill it in with a guess. A specification with no manufacturer confirmation is better left blank than invented — a mistake in material or dimensions leads to returns, not a faster sale.
If the supplier sends an updated price list regularly, note once which column maps to which field — work worth doing a single time.
Generate content in bulk, not by hand
Rewriting hundreds of listings by hand is the longest stage. Automation generates titles, descriptions and specifications from your templates in minutes — the job of the central catalog. Selba works with what you upload: it doesn't add properties the source data lacks, and it doesn't publish listings itself — you export the finished file and upload it yourself.
Next
Selba takes a price list, a CSV or a set of photos and turns them into finished listings that meet each marketplace's rules. Try it on your own catalog — the first 10 listings are free. The full list of platforms is in marketplaces and integrations.