ERP, order & WMS integration
Connect sales and purchase orders, products, customers, stock, warehouse work, shipment updates and financial hand-offs without duplicate entry.
Connect orders, products, stock, picking, packing, dispatch and finance across the systems your operation already depends on. ORBN designs, builds and supports vendor-neutral WMS integrations for UK warehouses and distributors.
Connect sales and purchase orders, products, customers, stock, warehouse work, shipment updates and financial hand-offs without duplicate entry.
Move marketplace, portal, supplier and customer documents through APIs, webhooks, EDI or governed file exchange with validation and traceability.
Add scanning, labels, catch weights, substitutions, proof and exception workflows where packaged WMS screens do not fit the operation.
Make failed, delayed or conflicting records visible; recover safely; reconcile stock and order states; and give every production flow an owner.
We map who owns each state, what warehouse staff actually do and how exceptions resolve before deciding whether the link needs an API, event or file.
Validation, duplicate protection, retry, quarantine, replay and reconciliation are designed with the happy path—not added after the first outage.
Representative orders, controlled production slices and parallel checks prove the records agree before a wider warehouse or channel moves across.
Score one representative warehouse flow from order acceptance to dispatch. The weakest authority, data or recovery assumption is usually more important than the average—and more useful than a long platform feature list.
Every slider starts at 1 for an unproven assumption. Score the real flow in scope—not the features claimed by either software vendor.
The largest risks are currently in ownership, warehouse rules or recovery rather than code. Start with one representative order flow, its exceptions and the records each system may change.
This scorecard supports initial scoping. It is not an interface audit, warehouse safety review or guarantee that a vendor exposes the required data.
ORBN is a vendor-neutral software and integration partner, not a WMS reseller. We help distributors, wholesalers, ecommerce operations and logistics teams retain platforms that fit, close the gaps between them and build only the workflow or integration capability the operation cannot buy sensibly.
Synchronise products, customers, purchase and sales orders, transfers, allocation, warehouse progress, shipment confirmations, inventory adjustments and finance hand-offs with explicit authority.
Route orders from B2B portals, Shopify, WooCommerce, marketplaces or other channels into the right commercial and warehouse flow, then return availability, fulfilment and tracking status.
Validate and trace purchase orders, acknowledgements, advanced shipping notices, invoices and partner-specific documents across EDI, managed transfer, APIs and structured files.
Connect parcel, pallet, route planning, vehicle, proof-of-delivery and customer notification services so labels, consignments, loads and tracking share the right shipment record.
Build focused mobile, barcode, label, catch-weight, substitution, quality and exception experiences around the WMS when the standard workflow creates manual work or unsafe shortcuts.
Expose delayed, rejected and conflicting records; give operations a safe correction path; reconcile material states; and assign technical and business ownership for production support.
Integration fails when both products appear authoritative for the same fact. A practical design assigns ownership at record and state level, then defines which changes may cross the boundary, how quickly and with what evidence.
Commercial order, customer, price, credit, tax, purchasing and invoice state commonly remain with ERP or OMS. Release, hold, cancel and change permissions must be explicit after warehouse work begins.
Warehouse allocation, work, location, pick, short, substitution, pack, container and dispatch evidence commonly belong to the WMS once a valid order is released.
A named master owns SKU identity, descriptions, barcodes, dimensions, units, cases, weights, batches and handling rules. Mappings are governed data, not code hidden in a connector.
On-hand, available, allocated, held, damaged, in-transit and customer-owned stock are different facts. The design states which system calculates each view and how adjustments reconcile.
The shipment, package, consignment, route, tracking reference and proof may span WMS, TMS and carrier services. Stable relationships prevent duplicate labels and ambiguous status.
Dashboards may combine records but should not become a hidden operational master. Definitions, refresh targets and lineage make warehouse and finance measures explainable.
Moving an accepted order into a warehouse is the easy demonstration. A production integration must keep later changes, shortages and shipment facts consistent enough for warehouse, sales, customer service and finance to act.
Confirm customer, address, warehouse, product, unit, quantity, date, service and hold state before creating warehouse demand. Invalid orders enter an owned exception queue rather than disappearing between products.
Return a durable warehouse reference and relate it to the commercial order and version. Repeated submission must not create a second order, and material rejection must be visible to the source.
Let the warehouse apply location, stock rotation, batch, ownership and wave rules. Decide whether allocation status returns to sales and how late changes are controlled after work is released.
Capture actual item, unit, quantity, batch, catch weight, substitute, short and reason. State who may approve a change and whether it updates the commercial order before packing.
Relate picked lines to packages, labels, pallets or containers; call carrier or routing services safely; and prevent repeated requests from creating chargeable duplicate consignments.
Publish shipped quantities, time, route or carrier, tracking and warehouse evidence. The ERP or OMS can then update customer status, documents, invoice eligibility and downstream commitments.
Handle cancellations, returns, failed delivery, write-offs, cycle counts and late corrections as named business events. Reconcile records at the level needed to explain a difference.
Microsoft’s current Warehouse Management guidance similarly separates master data, order documents and progress data, and describes shipment-order messages, business events and inventory alignment for connections to external ERP and order systems. See the official guide to exchanging data between warehouse systems. The exact objects differ by product, but the need to define authority and progress remains.
A warehouse estate rarely needs one mechanism everywhere. Supported product interfaces, operational timing and recovery needs should decide the method—not a blanket promise that every record will be “real time”.
Scroll horizontally to compare interface methods →
| Method | Useful when | Failure to design for | Minimum control |
|---|---|---|---|
| API request | A caller needs a response, validation or current record on demand | Timeouts, rate limits, partial failure and unsafe client retry | Stable identifiers, bounded timeout, duplicate protection and traceable errors |
| Event or webhook | Other systems must react when an order, pick or shipment changes | Duplicates, missing events, out-of-order delivery and hidden schema change | Versioned contract, idempotent consumer, replay and event-to-record correlation |
| EDI / managed file | Partners or established products exchange agreed business documents | Partner variants, delayed acknowledgement and files accepted but not processed | Document validation, control totals, acknowledgements and business-level status |
| Scheduled extract | Freshness can be measured in minutes or hours and bulk transfer is efficient | Overlapping runs, incomplete files, stale data and whole-batch replay | Watermarks, manifests, atomic hand-off, reconciliation and repeatable processing |
| Vendor connector | The supported connector covers the real flow and operating requirements | Feature assumptions, opaque mapping, licence limits and unclear support ownership | Representative proof, documented gaps, monitoring access and an agreed escalation path |
| Legacy database interface | No supported interface exists and a controlled read or vendor-approved route is viable | Tight schema coupling, unsupported writes, upgrade failure and bypassed business rules | Read-only by default, isolation, contract tests, change detection and an exit plan |
Monitoring server uptime is not enough. Operations need to know whether a specific order, line, product, adjustment or shipment has arrived, been accepted and reached the state the next team expects.
Version schemas, reject impossible identifiers and states early, and distinguish technical acceptance from successful business processing.
Make retries safe, preserve stable business keys and handle repeated or out-of-sequence changes without duplicating orders, picks, labels or stock movements.
Retry only transient failures, isolate records needing intervention, retain context and provide a controlled replay path after correction.
Trace a commercial order through warehouse work, packages, shipment and downstream updates. Alert on business delay, rejection and drift—not raw log volume.
Compare expected and actual records by agreed dimensions and time window. Produce an owned exception list and prove that correction does not create a second inconsistency.
Use least-privilege identities, encrypt transfer, protect secrets and personal data, audit privileged correction, contract-test vendor upgrades and keep rollback viable.
Define what the warehouse can still receive, pick, pack or dispatch during delay; how work is captured; and how records are safely brought back into sync.
ORBN delivers in production-shaped slices. Each stage reduces a named uncertainty and leaves artefacts the client can operate, challenge or hand to another supplier.
Follow representative inbound or outbound work with warehouse, customer service, finance and technical owners. Capture actual exceptions, manual bridges and timing constraints.
Name systems of record, identifiers, states, permissions, freshness, volumes, acceptance and reconciliation. Confirm vendor interfaces, licences and environments.
Test products, units, customers, addresses, stock and order history for missing, conflicting and anomalous values before they become failed warehouse transactions.
Exercise the hardest supported method, volume, mapping and exception with realistic records. Include repeat, delay and failure rather than demonstrating only one clean order.
Release a bounded flow with validation, security, monitoring, replay, reconciliation and operational correction. Measure time, error and warehouse impact.
Run representative acceptance, load and recovery tests; reconcile parallel results; train support; document rollback; and confirm every open exception has an owner.
Add flows, sites, customers or channels around evidence. Review incidents, data drift, vendor change, transaction cost and business measures with accountable owners.
“Connect ERP to WMS” is not an estimable scope. Price and schedule become more useful when the proposal exposes the operating flow, third parties and production obligations behind the connector.
Systems, warehouses, channels, record types, state changes, volumes, freshness and exceptions included—and the items explicitly excluded from the first release.
System-of-record decisions, identifier and unit mappings, cleansing, migration, history and the business owners who approve rules and corrections.
Vendor documentation, environments, credentials, licences, transaction charges, rate limits, support response and changes that another supplier must deliver.
Representative scenarios, performance, duplicate and failure tests, stock and order reconciliation, security, warehouse sign-off and measurable cutover conditions.
Pilot boundary, parallel running, data freeze, rollback, degraded operation, staff preparation, launch support and who decides whether the next site or channel proceeds.
Code, infrastructure and data ownership; monitoring and support responsibility; service targets; documentation; cloud, EDI and vendor fees; maintenance and exit route.
A well-bounded proof can often produce useful evidence in weeks, but the production schedule depends on interfaces, data, warehouse variation and cutover risk. If the real decision is whether to replace a platform, connect it or build around a process gap, use the ERP buy, integrate or build guide before treating integration as the predetermined answer.
Crowbond Foodservice was manually planning more than 200 daily orders across paper, spreadsheets and disconnected systems. ORBN built Crowlite around the real planning operation, integrating orders, routing constraints and driver workflows rather than forcing staff through a generic planning screen.
Daily orders coordinated through the planning workflow
Planning time reduced from about eight hours
Reduction in overall delivery cost
Reported annual saving from the operational change
These are results from one client engagement, not a forecast for every WMS integration. Read the evidence, scope and client account in the Crowbond route optimisation case study.
WMS integration connects a warehouse management system to the ERP, order management, ecommerce, EDI, carrier, automation, finance and reporting systems around it. The work includes more than moving fields. It defines which system owns products, orders, inventory, warehouse work and shipment status; translates identifiers and units; handles business events and exceptions; and provides monitoring, recovery and reconciliation when records do not agree.
A common division is for the ERP or order system to own commercial orders, customers and finance while the WMS owns warehouse execution such as allocation, picking, packing and dispatch. Product and order data flows into the WMS, then progress, inventory adjustments and shipment confirmations flow back. The interface may use APIs, events or webhooks, EDI, scheduled files or a supported vendor connector. The right design depends on source authority, timing, transaction volume, failure behaviour and each product’s supported interfaces.
Yes, subject to technical access and supplier constraints. ORBN can work with supported APIs, webhooks, databases, message queues and structured file exchange, and can place a controlled integration service around a legacy interface. We first verify documentation, licences, rate limits, data access, environments and vendor responsibility. Direct database integration is treated cautiously because unsupported writes or schema changes can corrupt data or break during upgrades.
Not every flow does. Order release, cancellation, priority changes and shipment status may need low latency, while catalogues, costs or analytical extracts may be safely batched. Real time adds failure, ordering, load and operational considerations. We define a freshness target for each flow and choose the simplest supported mechanism that meets the operational need, including what staff should do if the link is delayed.
The design gives records stable business identifiers and makes repeated delivery safe through idempotency or duplicate checks. Messages are validated before processing, changes are traceable, retries are controlled, and irrecoverable records are quarantined rather than silently dropped. Stock needs an explicit authority and adjustment model, plus scheduled reconciliation at the useful level of item, warehouse, location, status, batch, ownership or unit—not merely a comparison of grand totals.
A bounded flow between well-documented products can be proved in weeks. A wider rollout takes longer when master data is inconsistent, warehouse exceptions are undocumented, vendor environments are unavailable, several channels or sites differ, or cutover needs historical migration and parallel running. ORBN plans around production slices: prove the hardest interface and exception first, then expand by flow, warehouse, customer or channel with measurable acceptance conditions.
Cost depends on the number and quality of interfaces, records and transformations, order volume, latency, warehouse variants, exception paths, security, environments, migration, testing, cutover and support obligations. Licence or transaction charges from WMS, ERP, EDI and integration-platform vendors are separate. A credible proposal identifies assumptions, third-party dependencies, discovery and proof scope, production controls, acceptance measures and recurring operating cost rather than quoting only connector development.
Yes. Support can include monitoring, incident response, reconciliation, replay, defect fixes, vendor upgrade testing, performance and cost review, new channels or warehouses, and continuous improvement. The model should name first-line operational support, technical escalation, supplier responsibilities, service targets, access and runbooks. ORBN can operate the integration or hand over code, infrastructure, documentation and procedures to an internal or nominated team.
Map ordering, stock, warehouse, delivery, finance and customer workflows across the full distributor operation.
R/02Compare replacement, integration and targeted custom software against operational fit, control and whole-life cost.
R/03Design broader API, event, EDI and legacy interfaces with production monitoring, recovery and ownership.
R/04Build the warehouse, portal, mobile or operational workflow that packaged systems cannot support cleanly.
R/05Define trusted warehouse and commercial measures, data ownership, pipelines and operational decision products.
R/06See how one distributor reduced delivery planning time and cost through integrated operational software.
Bring the systems, the workflow and the exception your team reconciles by hand. We’ll identify the authoritative records, hardest proof and smallest safe integration scope.