Order-to-invoice flow across commerce, ERP and logistics
An illustrative integration programme: nightly exports and manual reconciliation replaced by a monitored, queue-based flow between a commerce platform, SAP Business One and a logistics provider.
By IT NET Solution
- Sector
- Wholesale and e-commerce (illustrative)
- Scope
- Order orchestration, stock synchronisation, invoice handover
- Platforms
- Sapier, SAP Business One, commerce platform, carrier API
- Engagement
- Architecture, build and managed operations
The challenge
Orders arrived from a commerce platform, a marketplaces feed and a small inside-sales team. They reached the ERP through a nightly file export, and stock figures on the website lagged behind the warehouse by up to a day.
When an export failed, nobody noticed until customers asked where their order was. Because the flow had no record of what had already been processed, the safe recovery was to re-check everything manually.
The approach
- 01
Map the process, not the endpoints
The team documented order states from capture to invoice, including cancellations, partial shipments and credit notes, before choosing any integration pattern.
- 02
Assign systems of record
Commerce owns the customer's checkout data, ERP owns pricing, stock and invoices, and the carrier owns tracking. Conflicts resolve in that order.
- 03
Make every step idempotent
Each message carries a business key, so a retried delivery cannot create a second order, a second shipment or a duplicate invoice.
- 04
Operate what was built
Queue depth, failed messages and reconciliation counts were placed on one dashboard, with an agreed response for each alert.
The solution
Sapier now orchestrates the flow as a sequence of monitored steps. A captured order is validated against ERP pricing, reserved against current stock and released to the warehouse; the carrier callback updates shipment state, and the invoice reference returns to the commerce platform for customer communication.
Failed messages land in a dead-letter queue with the original payload and the reason for rejection, so operations can correct and replay a single order instead of re-running a night of processing.
What changed operationally
- Order state is visible end to end, with the failing step identified rather than inferred.
- Stock figures move with the warehouse instead of the export schedule.
- Finance reconciles exceptions rather than reconstructing the month.
- New sales channels are added as documented adapters to the same flow.
The architectural point is not the platform. It is that the flow was designed for its failure cases first: duplication, ordering, partial success and recovery were decided before the first connector was written.
Topics
- Enterprise integration
- Sapier
- Order orchestration
- Operations
Related reading
Designing integrations for the failure case
The happy path is the easy part. Retries, idempotency, dead-letter queues and ownership are what decide whether an integration survives its first bad night.
Decide master data ownership before you connect systems
Most integration conflicts are ownership conflicts. A short method for assigning the system of record for customers, products, prices and stock before the first connector is built.
One customer record across sales, service and commerce
An illustrative retail programme: three CRM instances, two support inboxes and a commerce platform consolidated into a single customer record that sales and service both trust.
Working on something similar?
Describe the systems, the constraints and what needs to change. We will outline a realistic path and the trade-offs behind it.