SAP Business One and n8n: A Sales Reporting Example From Practice

Every morning, someone in sales exports the previous day's orders from SAP Business One, copies the figures into Excel, and builds a report for management from them. This costs half an hour before the first coffee, it is error-prone, and the figures aren't available until the day is already well underway.

SAP Business One and n8n can be connected so that this report is created without a manual export: n8n reads the new orders via the Service Layer interface of SAP Business One, consolidates them, and automatically delivers the result wherever it is needed. This article shows that using a publicly documented practical case, not a marketing promise.

What connects SAP Business One with n8n?

SAP Business One and n8n communicate via the Service Layer, the REST interface of SAP Business One that follows the OData standard. Through it, sales orders, business partners, or documents can be queried and written via HTTP calls.

The Service Layer has been a fixed part of SAP Business One since version 10.0. With Feature Package 2405, OData version 4 is the primary protocol, the base path changes from /b1s/v1/ to /b1s/v2/, and since version 10.0 FP 2305 the interface also supports OAuth 2.0 (Source: SAP Help Portal).

n8n is an automation tool based on the fair-code model, founded in Berlin in 2019, which can be self-hosted and, according to the provider, counts around 1.7 million active developers monthly. It connects over a thousand systems with each other, including SAP (Source: The Next Web).

That this combination is no niche experiment is shown by a report from May 2026: SAP invested around 60 million US dollars in n8n, at a valuation of 5.2 billion US dollars, and is embedding the tool into its own Joule Studio environment (Source: Yahoo Finance).

What does the documented practical case look like?

A practical case published in the SAP Community in October 2025 connects SAP Business One via the Service Layer with Google Sheets and a Telegram bot, generating a daily sales report from it without a manual export (Source: SAP Community).

The process consists of a few clearly defined steps:

• Reading: An OData query to the Service Layer retrieves the new sales orders with fields such as document number, document date, and business partner.
• Duplicate protection: A code step compares the order numbers with the rows already recorded and only takes over the genuinely new ones. This way, no duplicates arise when the process runs multiple times a day.
• Consolidating: A further code step calculates daily totals, determines the top customers, and the monthly average.
• Drafting: An AI step turns these figures into a readable report with a brief summary and notes on anomalies.
• Delivering: A Telegram bot sends the finished report to the relevant group.

The author of the article explicitly points out that the same pattern can be applied to purchasing, inventory, or CRM. Reporting is just the first use case, not the limit.

What building blocks does automated sales reporting need?

Robust sales reporting from SAP Business One needs five building blocks, and none of them is the AI itself.

• Data source: read access to the Service Layer, cleanly authenticated and limited to the required fields.
• State memory: a place that keeps track of what has already been reported, so nothing appears twice.
• Calculating: the aggregation of the figures in a code step, not by the AI. Sums and averages belong in traceable code, not in a language model.
• Drafting: the AI takes over the wording, i.e. what it is good at, and not the arithmetic, where it could miscalculate.
• Delivering: the channel through which the report arrives, be it Telegram, Microsoft Teams, or a simple email.

What does this mean for a mid-sized company with SAP Business One?

For a business with SAP Business One, this means: a recurring task like the morning report can be automated with manageable effort, without touching the ERP itself or purchasing an expensive additional license.

To be honest, though: such a workflow isn't built in an hour. Access to the Service Layer must be set up, the field names in the specific system checked, and special cases handled, such as canceled orders or documents with multiple currencies. The value is created not by the AI, but by the reliable mechanics behind it.

In our projects, we usually self-host n8n for this purpose, so that order and customer data never leaves the company's own network. That is exactly what distinguishes a controlled AI workflow from a black box: you see every step, and the data stays where it belongs. If you are planning something like this, our overview will help you with what to look for when hiring an n8n partner. And if the daily report is to become an acting AI agent, we'll show you where the sensible boundary between automation and a true agent lies.

Frequently asked questions about SAP Business One and n8n

Does SAP Business One need an additional license to talk to n8n?

No. The Service Layer has been part of the standard since SAP Business One 10.0. n8n accesses it for reading and writing via this REST interface; a separate integration license is not required for this. Costs arise mainly from setting up and operating the workflow.

Does the AI calculate the sales figures itself?

No, and that is intentional. Sums, averages, and top customers are calculated in a traceable code step. The AI only formulates the readable report from them. This way, calculation errors by a language model can be ruled out from the start.

Through which channels can the report be delivered?

The documented case uses a Telegram bot. n8n can just as easily send the report to Microsoft Teams, by email, or into a dashboard. The delivery channel is the most easily interchangeable part of the workflow.

Is this GDPR-compliant with German customer data?

Yes, if n8n is self-hosted. Then the order and customer data from SAP Business One never leaves your own network. It is important to choose the AI step so that it matches your own data protection requirements.