SAP integration
SAP remains the system of record for material master, pricing, and logistics. CataZenta is the system of engagement for rich product content, channel-specific copy, completeness, and marketplace syndication.
When SAP + CataZenta
| Scenario | Outcome |
|---|---|
| SAP has codes and weights; marketing owns copy | ERP nightly import → enrich in PIM → export to Amazon/Shopify |
| Multiple SAP plants, one brand catalog | Map plants to channels or locales |
| Long SAP descriptions, sparse attributes | AI attribute extraction after import |
| Listing errors on marketplaces | Fix in PIM with channel completeness — not in SAP views alone |
Reference architecture
Data mapping (typical)
| SAP concept | CataZenta target | Notes |
|---|---|---|
| Material number | Product code | Stable key for delta imports |
| Material type | Product family | One family per type (e.g. FERT, HALB) |
| Basic text / long text | Locale attributes | Split per en_US, de_DE, etc. |
| Gross weight / dimensions | Logistics attribute group | Unit of measure alignment |
| EAN / GTIN | Identifier attributes | Required for many channels |
| Plant / sales org | Channel or locale scope | Do not duplicate SKU rows unnecessarily |
Map once in a mapping document — reuse for IDoc, API, and CSV paths.
Integration patterns
1. Nightly batch (most common)
- SAP exports changed materials (IDoc, OData, or flat file to S3/SFTP).
- Middleware or CataZenta import job creates/updates products.
- Merchandisers enrich and pass completeness.
- Export jobs push to channels — export tutorial.
Delta key: changed_on or material status — avoid full reloads above ~50k SKUs without staging.
2. Event-driven (PLM + SAP)
New material released in SAP → message queue → import job for single SKU → workflow approval → syndication.
Use when SKU count is moderate and time-to-publish matters.
3. Enrichment back to SAP (optional)
Some enterprises push approved marketing text or GTIN corrections back to SAP via outbound IDoc. Define ownership: SAP wins on codes; PIM wins on customer-facing copy unless regulated otherwise.
Implementation checklist
| Phase | Tasks |
|---|---|
| Week 1 | Inventory SAP views; pick pilot material type; define family |
| Week 2 | Map 15–30 attributes; test 50 materials in staging tenant |
| Week 3 | Schedule import job; validate completeness per channel |
| Week 4 | First Amazon/Shopify export; fix job errors per SKU |
| Week 5+ | AI enrichment in review; MCP/API automation optional |
Operations
| Topic | Practice |
|---|---|
| Failures | Import job line items show per-material errors — fix mapping, not silent skip |
| Duplicates | Match on material number; never auto-merge different materials |
| Locales | SAP language keys → CataZenta locale tabs explicitly |
| Media | Images usually not in SAP — use DAM + link assets |
APIs and jobs
- Import / export jobs
- Product API
- Create product for manual pilot SKUs
Combine with ERP overview for pattern comparison (Odoo, Dynamics, etc.).
AI and agents
After SAP feed lands in structured families:
- Zen AI for copy and completeness questions
- MCP for “list materials imported last night below 80% completeness”
Partner and enterprise delivery
SAP landscapes vary (ECC, S/4HANA, CPI, PI/PO). CataZenta solutions teams provide mapping workshops and connector validation.
Book a demo — mention SAP version and material types.