Case Study:
How We Launched AFS Connector
Helping vendors migrate from Amazon Vendor Central to Amazon Seller Central and automate product data delivery through Amazon Feed Specifications (AFS)
Case Study:
Helping vendors migrate from Amazon Vendor Central to Amazon Seller Central and automate product data delivery through Amazon Feed Specifications (AFS)
In July 2018, Amazon introduced AFS (Amazon Feed Specifications) – a unified, JSON-schema-based standard for exchanging product data. Before AFS, vendors manually filled NIS Excel templates, which slowed down launches and introduced validation errors. As Amazon accelerates the shift toward Seller Central, vendors need a scalable, automated way to transform, validate, and push content that keeps pace with Amazon’s evolving taxonomy and marketplace rules.
Create a seamless connector that transforms vendor and Icecat content into AFS-compliant feeds and delivers them to Amazon automatically
Receive vendor product lists and pull enriched specs from Icecat.
Consolidate vendor extras with Icecat content into a unified model.
Align features, categories, units, and values to Amazon’s AFS JSON schema.
Run automated checks against AFS rules and marketplace-specific schemas.
Send feeds via Amazon APIs; parse delivery/validation reports.
Auto-handle taxonomy changes and re-map deltas without hard coding.
Misalignment between Icecat/vendor taxonomies and Amazon’s AFS
Frequent Amazon schema updates
Multi-country schema variations
Limited visibility into submission errors and validation
High manual effort for mapping and resubmissions
We mapped Icecat and vendor structures into Amazon’s AFS model and maintained rule sets that adapt to changes.
We integrated with Amazon endpoints to deliver feeds, retrieve validation reports, and auto-triage issues.
We maintained per-country JSON definitions, ensuring correct attributes and allowed values for each market.
We added monitoring and logs around submissions, validations, rejections, and deltas for transparent operations.
We exposed a UI to adjust mappings and attributes without hard-coding, cutting turnaround time dramatically.
Business Value: Major reduction in manual work; Gepard becomes the central source of truth.
We configured automated ingestion and transformation so Icecat and vendor content instantly aligns to AFS.
Business Value: Faster alignment with Amazon’s dynamic taxonomy and better data quality.
Strict mapping and validation workflows ensure feeds comply before submission.
Business Value: Consistent launches across geographies with localized schemas.
We manage schema differences per country so vendors can scale without rework.
Business Value: Faster issue resolution and continuous reliability.
We process Amazon’s validation reports automatically and surface actionable feedback.
Business Value: Fewer developer hours and quicker changes.
No hard coding; mapping changes happen in UI and propagate to future feeds.
Gepard PIM solution is a 100% flexible blend of product information management services, such as data validation, standardization, syndication, automated product information collection, data enrichment tools, and digital shelf analytics. With its extensive any-to-any connections base, the Gepard PIM platform allows you to integrate with various marketplaces, content providers, and retailers and build strong eCommerce partnerships, ruled by the power of top-notch products data.
By implementing a Gepard Syndicator within an AFS SaaS project, vendors speed up product data transportation and transformation, applying all necessary product feed changes automatically on a regular basis. Vendors enhanced product feed overall quality, having an opportunity to have their own data being merged with a CSPs (Catalog Service Provider) one.
