Visma and SAP Ariba integration

What moves between them

Purchase orders and supplier invoices flow from Visma into SAP Ariba. When a supplier invoice is created or modified in Visma, or when a purchase order is updated, ml-connector retrieves the record and submits it into Ariba for invoice matching and procurement reconciliation. Suppliers are synced in both directions so Ariba's supplier master reflects Visma suppliers. The sync runs on a daily or custom schedule based on your procurement cycle. Because Ariba's REST APIs are primarily read-oriented for procurement documents, most write operations go through Ariba's SOAP/cXML endpoints if bulk invoice creation or PO write-back is needed.

How ml-connector handles it

ml-connector stores Visma and Ariba credentials encrypted and manages two different OAuth flows: Visma's client credentials grant against https://connect.visma.com/connect/token with tenant_id in the request body and ipp-company-id on every API call, and Ariba's client credentials grant against https://api.ariba.com/v2/oauth/token with the static apiKey header appended to every call. Since Ariba offers no outbound webhooks, ml-connector polls Visma for new purchase orders and supplier invoices using the lastModifiedDateTime parameter on a schedule you define, then submits those records to Ariba's async reporting pipeline or SOAP endpoints. Suppliers are aligned in both directions so procurement records reference supplier keys that exist in both systems. ml-connector respects Ariba's rate limits and one-year async job window constraint by submitting multi-year backfills in annual chunks. Both tokens are cached and refreshed before expiry, and every record carries a full audit trail so failed submissions can be replayed once the downstream issue is resolved.

A real-world example

A mid-sized Nordic manufacturing and distribution company runs Visma.net ERP for accounting and purchasing across three warehouses and uses SAP Ariba as their centralized procurement portal for supplier sourcing, contract management, and invoice matching. Before the integration, the procurement team entered purchase orders into Ariba manually after creating them in Visma, and accounts payable manually matched supplier invoices in both systems, leading to duplicate work and reconciliation delays. With Visma and SAP Ariba connected, purchase orders flow automatically from Visma into Ariba when they are created or modified, and supplier invoices sync daily for invoice matching and three-way reconciliation. The procurement team no longer re-enters PO data, and the AP team starts each week with invoices already staged in Ariba for matching.

What you can do

• •Sync purchase orders from Visma to SAP Ariba on a schedule you define, eliminating manual PO entry into procurement.
• •Move supplier invoices from Visma into Ariba for invoice matching and three-way reconciliation without re-keying.
• •Keep supplier master data aligned in both directions so procurement records reference suppliers that exist in both systems.
• •Authenticate both Visma via OAuth 2.0 with tenant and company headers and Ariba via OAuth 2.0 plus API key on every request.
• •Poll Visma for changes on your procurement cycle with automatic retries, token refresh, and a complete audit trail for every sync.

Questions

Which direction does data move between Visma and SAP Ariba?

The primary flow is Visma into SAP Ariba. Purchase orders and supplier invoices move from Visma into Ariba for procurement processing and invoice matching. Suppliers are aligned in both directions so Ariba procurement records reference suppliers that already exist in Visma. Write operations for invoices and PO write-back to Ariba use SOAP/cXML endpoints rather than REST.

Does Visma webhooks support mean ml-connector can push data instantly?

Visma webhooks require explicit enablement at the company level and deliver events once with no automatic retry, so they are unreliable for critical integrations. ml-connector uses webhook events where they are enabled but implements polling as a fallback and primary mechanism, querying Visma for changes at a schedule you define to ensure no records are missed.

How does ml-connector handle Ariba's async reporting jobs and one-year window limit?

When syncing historical data spanning multiple years, ml-connector submits separate async reporting jobs in annual chunks within the one-year constraint, polls each job for completion via jobId, and downloads results in 50,000-record pages. New data is synced daily in smaller date ranges that fit within limits, and tokens are cached and refreshed before expiry to avoid authentication failures.