SAP Salesforce integration

What a SAP to Salesforce integration actually does

SAP is the system of record for what a customer costs, owes, and has actually ordered. Salesforce is where sales works: pipeline, quotes, relationships, forecasts. The two describe the same customers and the same deals from opposite ends, and for most companies they stay in sync because somebody re-types the information from one screen into the other.

A SAP to Salesforce integration removes that person. It pushes the ERP truth - who the customer is, what they may buy, at what price, what they still owe - into Salesforce so sales works from real data, and it writes closed deals back into SAP as sales orders so fulfilment and finance start without a re-key. The point is not to make the two systems identical. It is to let each own what it is good at and stay honest with the other automatically.

What data moves

SAP object / event	Becomes in Salesforce	Notes
Business Partner / customer master (KNA1)	Account	Matched by SAP customer number to an external ID field; upserted, never duplicated
Contact persons	Contact	Linked to the parent Account; partner-function role preserved
Material master	Product2	Only the sellable materials for the relevant sales org / distribution channel
Condition-based pricing	PricebookEntry	SAP is the pricing authority; currency and scale prices carried across
Sales order status & delivery	Order / status field	Write-back of confirmation, delivery, and shipment status against the originating deal
Billing document / invoice	Order or related record	So sales sees what was actually invoiced, not just what was quoted
Open items / credit exposure (AR)	Account fields	Drives credit warnings and can block a Salesforce Order before it reaches SAP

The reverse flow matters just as much: a closed-won Salesforce opportunity or Order becomes a SAP sales order, with partner functions, sales org, and materials resolved on the way in. Which objects move, and in which direction, is agreed once and encoded in the pipeline.

The details that break naive syncs

A point connector or a nightly export gets the happy path working and then leaves the hard 20% on your desk:

• ID matching and dedup. SAP identifies a customer by KUNNR or Business Partner ID; Salesforce by its own record Id. Without a dedicated external ID field and upsert logic, sales-created Accounts and ERP-created customers become two records for the same company. The mapping, not the copy, is the integration.
• Org structure has no CRM equivalent. SAP sales org, distribution channel, and division determine which materials and prices are even valid. Salesforce has no native concept of them, so the pipeline has to hold that context and filter accordingly - not push every material to every Account.
• Partner functions. A SAP sales order needs sold-to, ship-to, and bill-to (AG / WE / RE) resolved separately. A Salesforce opportunity usually carries one Account. Mapping one to three correctly is where order write-back quietly fails.
• Pricing and currency. Prices are computed in SAP from condition records, in a specific document currency, with SAP’s rounding rules and decimal handling (a zero-decimal currency like JPY is not a cent currency). Recomputing in Salesforce guarantees drift. We carry SAP’s numbers across.
• Delta versus full sync. Reloading every customer nightly does not scale and burns Salesforce API limits. We drive deltas from SAP change pointers or S/4HANA CDS delta tokens and Salesforce SystemModstamp, so only what changed moves.
• Partial failures and locking. One order in a batch failing SAP validation - credit block, missing material - must not fail the batch. Parallel updates to the same Salesforce parent Account can throw row-lock errors that need sequencing, not blind retries.

How we build and run it

We treat this as a pipeline, not a scheduled export. Objects are read through the right SAP interface per case - OData and CDS views on S/4HANA, IDoc or RFC/BAPI through SAP Gateway on ECC - validated, mapped to your agreed Salesforce model, and written with Bulk API 2.0 for volume or targeted upserts for deltas.

The pipeline is idempotent. Every record carries a stable key - the SAP customer or document number on one side, the Salesforce external ID on the other - so a retry or a re-run updates in place and never creates a duplicate. It runs on cloud-native, fully EU-hosted AWS infrastructure, so customer and order data never leaves the EU, which keeps your DPA / AVV and GDPR position clean.

Then we keep it running. Monitoring, alerting, incident response, and watching for SAP and Salesforce API changes - a new S/4HANA release, a Salesforce API version sunset, a governor-limit change - are our responsibility under contract. You get a named owner and an SLA, not a script somebody has to remember to run.

When this integration is worth building

If a handful of accounts move between the systems and sales is happy re-keying the occasional order, a manual process or a simple point connector is genuinely fine, and we will tell you so. The integration earns its place when order volume climbs, when order-to-cash write-back has to be reliable, when credit exposure needs to reach sales before a deal is booked, or when duplicate Accounts and drifting prices are already costing your team more time than a managed pipeline would.