Business Central integrations break down when each external system gets its own custom wrapper, exception handler, and field map. The first API script may look inexpensive. The tenth one becomes a maintenance portfolio.
Component breadth is what keeps the tenth endpoint cheap. COZYROC SSIS+ includes and , built across market. Licensing stays fixed at for an Ultimate subscription with 2 license keys, so each new Business Central endpoint reuses the platform instead of justifying another wrapper service.
Why Business Central API Work Becomes Expensive
Dynamics 365 Business Central often connects to e-commerce stores, warehouse systems, payroll providers, tax services, logistics networks, and reporting platforms. These systems communicate through REST and OData patterns, which are useful but still require careful work:
- Authentication and environment configuration.
- Endpoint-specific request and response handling.
- JSON and OData metadata interpretation.
- Error routing when one order, invoice, or item update fails.
- Mapping updates when analysts add custom fields or extensions.
The gives Business Central teams a familiar SSIS design surface for this work. REST Source, REST Destination, the REST Connection Manager, and DFT+ let developers model the integration as a package instead of building a separate wrapper service for every endpoint.
COZYROC Components For Business Central
| Pain point | COZYROC component | Architectural result |
| Business Central endpoints need repeatable REST/OData configuration | REST Connection Manager and OData REST configuration | Endpoint, authentication, resource, and field definitions can be managed as package configuration |
| Order, item, customer, and invoice data must be read from APIs | REST Source | API responses are converted into SSIS rows for validation, lookup, and staging |
| External records need to create or update Business Central objects | REST Destination | Writes are mapped in SSIS data flow instead of hand-built JSON payload scripts |
| Field additions or extension changes break static packages | DFT+ / Dynamic Data Flow Task | Runtime metadata mapping reduces redesign work when schemas evolve |
| A single bad row can interrupt a batch | REST Destination error routing and package-level validation | Failed rows can be isolated for remediation instead of hidden inside custom wrapper code |
| Empty or unchanged values can overwrite valid ERP data | SSIS Conditional Split or Derived Column transformation before REST Destination | Payload can be filtered before submission so only fields that carry updated values reach the Business Central endpoint |
Order-To-Cash Implementation Pattern
A practical Business Central integration should not write directly from a storefront into the ERP without a control layer. Use this package pattern:
- Pull orders, customers, items, shipments, and payment references through REST Source.
- Stage raw input in SQL Server or another controlled landing table.
- Run validation against item numbers, customer accounts, tax codes, dimensions, and ship-to records.
- Use DFT+ for metadata-driven mapping where custom fields are expected.
- Route valid rows to REST Destination for Business Central writeback.
- Route rejected rows to an error table with endpoint, request payload, response code, and remediation owner.
When the package is promoted to , operations can schedule the load with or trigger it through when an external system posts a new event.
TCO And ROI Framing
Business Central integration cost is usually dominated by developer hours after go-live. The ROI case for COZYROC is strongest where the same pattern repeats across many endpoints:
- One package pattern can serve multiple entities.
- REST and OData handling are configured in the package layer.
- DFT+ reduces repair work after custom-field changes.
- A fixed SSIS+ platform cost is easier to plan than a growing set of wrapper services and variable usage fees.
FAQ
Can we connect Business Central OData endpoints without writing wrapper services?
Yes, the handles Business Central OData v4 endpoints inside SSIS. Connection and authentication configuration are maintained in the package, and REST Source converts API responses into typed SSIS rows for validation, lookup, and staging. The integration is modeled as a Visual Studio package a future maintainer can open and modify, rather than a separate service with its own deployment and exception handling.
How do we stop one bad order from failing an entire batch load?
Route failed rows instead of failing the run. REST Destination error routing and package-level validation isolate bad records into an error table with endpoint, request payload, response code, and remediation owner. The rest of the batch posts normally, and the rejected order becomes a work item with enough context to fix, not a hidden exception inside wrapper code.
How do we avoid overwriting valid Business Central data with empty values?
Add a Conditional Split or Derived Column transformation before REST Destination to exclude null and unchanged values from the payload before it reaches the Business Central endpoint. Business Central’s OData v4 API processes update operations field by field, so limiting the payload to only fields with valid new values prevents accidental overwrites of existing ERP data. Pair this with a staging layer that validates item numbers, customer accounts, tax codes, and dimensions before any row reaches REST Destination.
Can an external system start a Business Central load the moment an event happens?
Yes, when the package runs in , can trigger execution when an external system posts a new event. Recurring loads run through Scheduled Jobs instead. That keeps trigger logic in the platform and data movement in the package, so the storefront never needs direct write access to Business Central.