Zum Inhalt springen
Retailer checking marketplace orders on a smartphone while packing a parcelAI-generated

E-Commerce

Online marketplaces: prepare and publish reliable product data

Online marketplaces bring together offers from different sellers on one platform. For manufacturers and retailers, they provide additional sales channels, each with its own requirements for product information, images and order handling.

Selling on several marketplaces therefore requires more than copying an item list. You need a traceable path from maintained source data to the right offer: choose a category, complete required attributes, assign variants and media, transfer the data and handle feedback. This guide explains how to plan that process with PIM, DAM and a suitable channel integration.

What is an online marketplace?

An online marketplace is a digital platform where multiple providers offer goods or services and buyers find relevant offers. The operator supplies the platform and sets its rules. Any additional services, such as payments or shipping, depend on the particular offering and agreement.

By contrast, a retailer or manufacturer operates its own online shop for its product range. A marketplace can extend reach, but also brings fees, competition and requirements. Using an existing platform does not remove the need to maintain product information or organise delivery, returns and service.

Types and selection: which marketplace fits your range?

Start by distinguishing the product range and audience. A horizontal marketplace covers different product categories; a vertical marketplace focuses on an industry or segment. B2B offerings serve business customers, while B2C offerings serve consumers. Whether the operator also sells its own goods or operates internationally are further characteristics. These can overlap.

Familiar names for initial research include Amazon, eBay and OTTO. A ranking alone will not tell you which platform suits your business. Before planning the integration, check:

  • Range and audience: are there suitable categories and buyers for your products?
  • Commercial viability: how do fees, shipping, returns and any advertising costs affect your calculations?
  • Operations: who handles orders, stock changes, customer questions and returns?
  • Data requirements: which attributes, variants, images and languages does the target channel require?
  • Integration: how will data be transferred, which systems are involved and who handles error messages?

Start with one product group and one target channel. A representative test tells you more than the number of marketplaces that can theoretically be reached.

Establish required attributes, categories and values first

The information required depends on the marketplace and product category. A free-text field in your PIM is not automatically a valid value in the destination system. Data types, units, permitted values and dependencies between fields can also matter.

The technical documentation illustrates this: Amazon’s Product Type Definitions API provides product type requirements as schemas. The eBay Taxonomy API describes category-specific attributes, including required information, allowed values and suitability for variations. Use the current requirements of the specific target channel when implementing the connection.

Create a data profile for each channel: which source fields supply which destination fields? Which transformations are allowed? What happens when a required value is missing? Assign missing information to the responsible team. Guessing a value just to get an export through does not improve the offer.

Use this overview to prepare:

Data area Establish in the source Check for the target channel
Identification Internal item number and existing product identifiers Which identifiers does the channel require or accept?
Category An appropriate classification for the product Which target category and mandatory attributes apply?
Attributes Material, colour, dimensions and other verified properties Data types, units and allowed values
Copy and language Approved titles, descriptions and translations Intended fields, language and content requirements
Variants Relationship between the model and sellable items Permitted variation attributes and grouping
Media Assigned, approved images and documents Order, format, image content and accessibility

Check variants and images for each item

A model is not always the sellable unit. A product in two colours and two sizes may result in four different items. Each needs an unambiguous association with its attributes, identifiers and media. Check separately which combinations the marketplace allows you to group as variations.

A simple planning example: a drinks bottle comes in blue and black, with capacities of 500 and 750 millilitres. The item “blue, 750 ml” must not receive the black bottle’s image or a capacity of “500 ml”. Its internal item number, relationship to the model and attributes must stay consistent during export. This example does not describe the requirements of a particular marketplace.

For media, check in particular:

  • Does the main image show the actual variant being offered?
  • Do detail images and documents match the same product version?
  • Are the image format, resolution and presentation suitable for the channel?
  • Can the recipient retrieve the image files provided?
  • When an image changes, is the new version transferred and displayed?

A DAM system supports central media management and links assets to products. The outputs and checks you automate need to match the specific channel integration.

Mapping: turn source data into a suitable offer

Mapping connects the fields and values in your source with the destination’s data model. Renaming columns is often not enough. An internal category must match a marketplace category, a colour name may need to map to a predefined value, and a measurement may need to be split into its value and unit.

Document the rules with examples. If the source contains “0.75 l” and the agreed destination profile expects a numeric value in millilitres, the conversion produces “750”. The unit must be unambiguous. Whether this format is accepted by a particular channel depends on its requirements.

Also establish whether an empty field means “leave unchanged”, “delete the value” or “invalid input”. This distinction matters especially for updates. Test unknown values and missing required fields before transferring a larger range.

Synchronise product data: distinguish transfer from outcome

A successfully sent file does not prove that an offer is displayed correctly. Plan the process through to feedback from the destination system and inspection of the published product presentation.

1. Establish data ownership

Define the authoritative system for each field. Descriptions and product attributes may be maintained in PIM, with media in DAM. Depending on your systems, prices, stock and order data may come from ERP, inventory management or other applications. A PIM system does not automatically replace inventory and order processing.

2. Prepare and check the export

Select the intended range, language and channel. Check completeness, values, variant relationships and media assignments before transfer. Record which version of the data was approved and which mapping rules apply.

3. Transfer and associate feedback

The handover may use files, APIs or an additional integration platform. Establish where transfer status and error reports will be visible. Each message needs to lead back to the affected item, field and processing run so that someone can act on it.

4. Correct and check again

Where possible, correct an inaccurate source value in the authoritative system. If the problem is in the mapping, adjust the rule and check other affected items. After resubmission, review both the feedback and selected offers in the target channel.

5. Test changes and removals

Alongside new products, test changed descriptions, replacement images, variant changes and the intended removal of an offer. Establish which transfers contain the full dataset and which contain changes only. Agree on update intervals and how interruptions will be handled. A successful export does not imply simultaneous publication on every platform.

How PIM, DAM and OMN work together

OMN PIM centrally manages product information and connects it to media. OMN DAM organises the associated assets. OMN Channel Management assembles content for different output channels. OMN EAI provides import and export profiles, together with the history and logs of the corresponding runs.

The marketplace connection is a separate part of the planning. The interface or additional platform required depends on the target channel and your existing systems. A documented example is BÄR Schuhe: product data and media are managed with OMN, while marketplaces are connected through Tradebyte. The success story names Amazon, Decathlon and Otto among the target channels. This does not imply a direct connection to every marketplace in every OMN project.

Checklist for your first marketplace test

Assemble a small test range containing both typical products and difficult cases. One fully maintained variant is not enough.

  1. Target categories, required attributes and allowed values are documented.
  2. Authoritative systems and responsible people are identified for each data area.
  3. Items, variants, images and languages are assigned unambiguously.
  4. Missing values and incorrect formats are detected and assigned for resolution.
  5. Feedback can be traced to the item and transfer.
  6. Corrections, resubmission, image replacement and removal of offers have been tested.
  7. Someone checks the actual presentation on the marketplace.

Then observe which errors recur, how much rework is needed and how long an approved change takes to become a verified offer. Use these findings to improve the process and add further product groups.

Common questions about online marketplaces

Do I need a PIM system to sell on a marketplace?

No. Manual maintenance may be sufficient for a small range. More variants, languages and target channels increase the coordination involved. What matters is whether you can reliably maintain, check and reuse product data.

Can I send the same product file to every marketplace?

Not without comparing the requirements. Categories, field names, allowed values and media specifications can differ. A central data source is useful; the output must fit each destination profile.

Does an accepted transfer guarantee a correct offer?

No. In addition to technical processing, you need to check that the right variants, descriptions and images appear in the destination channel. Feedback and spot checks of visible offers belong together.

Discuss your marketplace data process with apollon

Bring a typical product group, a target marketplace and examples of your existing data or recurring errors to a demo. This provides a concrete basis for discussing the maintenance, preparation and integration your team needs.

Request a demo