← All eCommerce integrations

The workflow at a glance

  1. Supplier XML / CSV / API
  2. Mapping and validation
  3. Pricing and availability rules
  4. Controlled catalogue update

What we need before implementation

  • Sample data and permission to use it
  • Stable supplier identifiers
  • Field ownership and pricing rules
1.

At a glance

A supplier file needs rules before it becomes a store catalogue. Integration identifies existing products, validates changes and updates only fields for which the supplier is an authorized source.

2.

Give every product a stable identity

Define a stable key and retain a mapping for each supplier. EAN is useful when valid and available, but not every product has a unique reliable barcode. Colour and size variations need explicit handling. If two suppliers offer the same item, decide whether one commercial record will have multiple supply sources or whether separate products are justified. Supplier-specific references must remain traceable without replacing the store’s internal identity.

3.

Separate data updates from commercial decisions

Map titles, attributes, categories, images, weight and cost. A supplier may update cost and availability while the business retains final descriptions and selling prices. Margins, tax treatment and rounding are defined explicitly. A sudden cost reduction or empty price must not be published without checks. Protect the team’s edited work against blind replacement. Record which fields belong to each source so a later import cannot silently overwrite another system’s authoritative values.

4.

Supplier availability is not store inventory

Supplier availability can change and may include a lead time. Separate it from the store’s physical stock and apply agreed display rules. If an item temporarily disappears from a file, do not immediately delete it. Distinguish full-catalogue snapshots from change-only feeds so absence is interpreted correctly. When multiple suppliers disagree, resolve availability through approved rules instead of summing quantities that may represent the same upstream stock.

5.

Batches, history and failure handling

Each import records its time, source, record count and changes. Validate data before updating the live catalogue. Failed records enter a separate queue, while successful ones are not recreated on retries. Test Greek text encoding, broken image links and size limits. An unusually small file or changed structure should stop publication and alert the owner. Small initial batches make it easier to inspect mappings and reverse incorrect field updates.

6.

An example using multiple sources

For an electrical-supplies retailer, physical inventory comes from the ERP and orderable quantities come from two suppliers. The integration preserves internal codes, selects permitted supply sources and displays agreed lead times. Acceptance testing covers new items, cost changes, a shared item from another supplier and temporarily empty files. The team receives a changes report rather than manually searching the catalogue. Reconciliation also shows records that could not be mapped confidently.

Acceptance checks before launch

  • Retries do not create duplicate products
  • Empty files do not delete the catalogue
  • Edited descriptions and ERP stock are protected

FROM READING TO ACTION

A clear plan for your platform.

Tell us where your platform struggles and what needs to improve. Our team can connect the diagnosis with code, performance and infrastructure work.

Discuss your platform

Sources and further reading