If you want to lose a major B2B customer in 2026, there is an easy way: keep forcing them to order by fax, email or phone. Professional procurement teams have long been working out of their ERP or e-procurement systems and expect suppliers to plug seamlessly into those workflows. The bridge between the buyer's procurement system and the supplier's shop is called punchout - and the two formats carrying that bridge are OCI and cXML. In this guide we show how to connect your B2B Shopware shop to SAP Ariba, Coupa, Onventis and other platforms - in a technically clean, GDPR-friendly and economically sensible way.
Why B2B buyers in 2026 do not order the old way anymore
The global B2B e-commerce market is growing to roughly 36.16 trillion USD in 2026 (Capital One Shopping Research), with a growth rate of 14.5 percent year-on-year. Germany alone reached 530 billion EUR in B2B e-commerce in 2024 (BEVH Impact Studie 2025) and is expected to cross 465 billion EUR in the DE B2B online segment by 2026 (Trade.gov). The comparison with B2C shows the dimension: the average B2B order value is 491 USD versus 147 USD for B2C (Shopify Enterprise 2025).
In parallel, buyer behavior has completely changed. The Sana Commerce 2025 study shows that 73 percent of B2B buyers now actively prefer to buy online (Sana 2025). Even more importantly: 75 percent consider switching suppliers when the digital systems are outdated or cumbersome (Sana). And the orders are no longer small: 39 percent of B2B buyers today place orders above 500,000 USD without sales reps, fully self-service (McKinsey B2B Pulse 2024), while Forrester predicts that in 2025 more than 50 percent of all transactions above 1 million USD will be self-serve (Forrester Predictions 2025).
- 10.2 channels are used on average by B2B decision makers during a purchase journey, with e-commerce being the largest revenue channel at 34 percent (McKinsey B2B Pulse 2024)
- 61 percent of B2B buyers prefer a rep-free buying process - without the detour through sales (Gartner 2025)
- 46 percent name real-time stock availability as a must-have, no longer a nice-to-have (Sana)
- One third (33 percent) of B2B orders placed online today contain errors - and 68 percent of buyers get pushed away from further online ordering because of it (Sana B2B Buyer Report 2025)
- Manual data entry in B2B has an error rate of 4.2 percent (Conexiom) - every line is a potential complaint
Our article on the B2B self-service portal with Shopware 6 describes these baseline expectations in detail. Punchout is the next evolution: while a self-service portal opens a direct path into your shop, punchout connects your shop seamlessly into the buyer's existing e-procurement system - so that the buyer does not even notice they are entering another system.
What punchout actually is (and what OCI/cXML stand for)
Technically, the term punchout describes a process where a buyer, from inside their own e-procurement system, does not click a static catalog but instead punches out live into the supplier's web shop - using it as an embedded catalog source. The buyer sees customer-specific pricing, current stock and their real assortment, fills a cart and sends that cart back to their own procurement system via a standardized protocol. There the cart goes through the internal approval and budget workflows before a final purchase order is actually sent to the supplier.
Behind the scenes, two dominant standards do the work: OCI (Open Catalog Interface) was originally developed by SAP and is the de facto standard for ERP-driven procurement in Europe and especially the DACH region. cXML (commerce eXtensible Markup Language) was pushed by Ariba and is leading in North America and on platforms like Coupa and Oracle Procurement Cloud (Greenwing Technology). Anyone running B2B internationally or across industries typically has to support both - a pure cXML or pure OCI stack is rarely enough.
The market is fragmented but not exotic: there are more than 75 relevant e-commerce systems and over 220 e-procurement solutions worldwide that are commonly connected via punchout (TradeCentric). According to Digital Commerce 360, e-procurement already accounts for 8.5 percent of all US digital commerce sales and 7.1 percent of all B2B sales (Digital Commerce 360). In Europe adoption is even more concentrated, since SAP leads with a 29.1 percent global market share in procurement software (Apps Run The World), and Germany represents roughly 40 percent of the European procurement software market (Global Market Insights).
OCI = HTML-form based data exchange, today mostly in version 4 or 5. Simple, quick to implement, SAP-driven. cXML = XML-based data exchange with strong schema definitions, developed by Ariba, widespread on Coupa, Oracle and SAP Business Network. Both formats solve the same use case - setup request, browse, PO return - just with different syntax and authentication.
OCI vs. cXML in direct comparison
Without understanding the differences, you cannot build a clean punchout strategy. The following table summarizes the real-world differences we see time and again in Shopware integrations:
| Aspect | OCI (Open Catalog Interface) | cXML (commerce eXtensible ML) |
|---|---|---|
| Origin | SAP (around 1998) | Ariba (around 1999) |
| Reach | Europe, DACH, SAP customers | North America, Coupa, Oracle, Ariba |
| Format | HTML form POST, URL parameters | Structured XML documents |
| Current version | OCI 4.0 / 5.0 | cXML 1.2.x (stable) |
| Authentication | HOOK_URL, user/password | SharedSecret, SenderCredential |
| Cart return | Form POST with key/value pairs | PunchOutOrderMessage (XML) |
| Implementation effort | Low to medium | Medium, schema validation |
| Typical platforms | SAP Ariba (Classic), SRM, JAGGAER, Onventis | Coupa, Oracle, SAP Business Network, Ariba |
The key practical difference: OCI transports cart data back through a classic HTML form, which is quick to implement but has limits with complex additional fields. cXML is structured, validatable and carries metadata such as delivery addresses, cost centers and tax data more cleanly - but in return needs proper schema handling and clear error management on both sides.
How a punchout transaction works technically
The flow shown above illustrates the four stations of a punchout process. In detail, every transaction runs through the following steps - regardless of whether OCI or cXML is used:
- Setup request: The buyer clicks the supplier catalog inside their e-procurement system (for example SAP Ariba, Coupa, Onventis). That system sends an authenticated request to the Shopware shop - for OCI as a form POST with HOOK_URL, for cXML as a PunchOutSetupRequest XML document.
- Authentication and session setup: The shop validates the credentials (SharedSecret for cXML, typically username/password for OCI), identifies the buyer, loads their customer group, customer pricing and assortment, and generates a one-time session token.
- User browses: The buyer is redirected (or embedded via iframe) into the Shopware shop. They see the B2B assortment with their individual terms - without noticing they just entered another system.
- Cart assembly: The buyer adds products to the cart like in any other shop. They can use quick order via SKU, volume pricing, budgets and all B2B features as long as these are enabled - including modern express checkout mechanics for recurring orders.
- PO return: Instead of triggering an order, the checkout button triggers a cart return - the cart data is sent back to the HOOK_URL as an HTML form POST (OCI) or as a PunchOutOrderMessage XML (cXML).
- Approval in the buyer system: Inside e-procurement the cart goes through the internal workflows - budget check, cost center assignment, manager approval. A real order only emerges after this process.
- OrderRequest back to the supplier: The final order is then sent back to the supplier again - typically as a cXML OrderRequest, an EDI document or directly via ERP integration. Only at this point does a sales order appear in the ERP system.
An important point to understand: in a punchout flow the buyer does not leave the shop with a sales order, but with a cart. The actual order may come back hours or days later via an OrderRequest. This has serious implications for inventory, price locking and ERP processes - and needs to be designed carefully.
Efficiency gains in numbers
Punchout is not an end in itself but delivers hard efficiency gains on both sides of the procurement chain. Key benchmarks from recent reports: A fully automated purchase order chain requires 24 hours instead of 35 hours cycle time - a reduction of 11 hours per transaction (APQC). Top performers approve 98 percent of their POs electronically while bottom performers are at around 50 percent (APQC).
The cost side is even clearer: Best-in-class organizations process a single invoice for 2.88 USD while non-automated processes cost 12.88 USD (Ardent Partners 2025). In Germany the processing time per invoice drops from 45 minutes to 3 minutes as soon as structured e-invoices come into play (Trade.gov). And the Hackett Group reports that automated P2P systems reduce the exception rate from 22 to 9 percent and cut deviation by 37 percent (Hackett Group).
-85% order cost
Punchout integration can cut order processing costs by up to 85 percent (Procurify)
-11 h cycle time
24 hours instead of 35 hours for the entire PO process (APQC)
-78% invoice cost
2.88 USD instead of 12.88 USD per invoice in best-in-class operations (Ardent Partners 2025)
-37% deviations
Automated P2P systems reduce exception rate from 22 to 9 percent (Hackett Group)
98% e-approval
Top performers approve nearly all POs electronically, bottom performers only 50 percent (APQC)
60% SMBs digital
In Germany 60 percent of SMBs have already digitized procurement processes (BME/Onventis 2024)
The German Bitkom 2025 study further shows that 95 percent of companies cite simplified ordering processes as the main benefit of their digitalization (Bitkom 2025). Still, adoption is only beginning: 60 percent of German SMBs have digitized procurement processes, another 30 percent plan to do so in the short term (BME/Onventis 2024). Punchout is therefore a clear competitive advantage for suppliers who ride this wave early.
Curbing maverick buying through catalog compliance
One of the biggest drivers for punchout projects on the buyer side is the containment of maverick buying - orders that bypass the central procurement process and therefore neither use the negotiated terms nor meet compliance requirements. The numbers are striking: Companies lose 5 to 16 percent of their savings targets to maverick buying (Ivalua/Sievo). On a 500 million USD spend base, that is 25 to 80 million USD lost - per year.
Even best-in-class organizations with established e-procurement systems only achieve around 69 percent contract compliance in practice (Aberdeen/Sievo). The industry target is maverick spend below 10 percent and PO compliance above 95 percent (Fairmarkit) - and without punchout catalogs that is simply not realistic to achieve in most organizations.
For suppliers, maverick buying is paradoxical: it sounds like more revenue when buyers bypass the system - but usually means losing a strategic supplier position. Once procurement discovers the pattern, the supplier either ends up in the compliance catalog or on the chopping block. A properly connected punchout catalog turns the supplier into a strategic partner instead of a risk source.
Technically the point is that the Shopware shop only shows products, prices and volume tiers released for that specific buyer. That requires a clean link to the B2B pricing strategies and a clear product whitelist mechanism. Without these, a punchout catalog becomes a compliance risk - with the right configuration it becomes an enforcement tool for both sides.
Punchout in Shopware 6: B2B Components instead of B2B Suite
With the announced Shopware 6.8 release in early 2026 Shopware is drawing a strategic line under the previous B2B Suite and ships its features as B2B Components directly in the core (Shopware). For B2B integrations, including punchout scenarios, this means: quick order, role model, approval, quote, shopping lists and budgets are available out of the box and form the natural basis for the punchout session.
This move fits Shopware's market position: in the Paradigm B2B Report 2025 Shopware received 11 out of 12 medals (Shopware) and recorded growth of plus 300 percent in North America during the first half of 2025 (Shopware/PR Newswire). The vendor clearly positions itself as a serious B2B alternative - no longer just a mid-market shop system.
In practice, punchout in Shopware 6 cannot simply be turned on via a plugin - it is a layered topic: the setup request must be authenticated, the buyer must be mapped to a Shopware customer group, their session embedded in a URL and the cart return delivered via a dedicated endpoint. All of this happens on top of the existing Shopware APIs and can be implemented cleanly based on the B2B Components.
Integration with SAP, Microsoft Dynamics and ERP
A punchout flow does not end with the cart return. For the order to actually become economically real, the OrderRequest must flow cleanly into the ERP - and bidirectionally: prices and customer data from ERP to shop, orders and returns back from shop to ERP. The most common systems in DACH are SAP Business One, Microsoft Dynamics 365 and JTL-Wawi for mid-size merchants.
Reality is not always pretty: ERP custom web services for SAP and MS Dynamics typically cause development costs in the four- to five-figure EUR range and implementation times of 3 to 6 months (Alumio). For Dynamics 365 Finance & Operations it is also true that customer-specific pricing, real-time stock and custom fields are not B2B-ready out of the box (Silk Commerce) - meaningful customization is practically always mandatory here.
On top of this is the mandatory e-invoice: since 2025, B2B invoices must gradually be exchanged in structured formats (XRechnung, ZUGFeRD) - and yet only 45 percent of German companies can currently receive structured e-invoices (Bitkom 2024/25). A punchout project is therefore often the right moment to rethink the entire supplier-buyer chain - from the catalog all the way to the DATEV interface in accounting.
Top e-procurement platforms in DACH
Anyone doing B2B in the German-speaking region will come across the same e-procurement platforms repeatedly. A short overview of the relevant systems - as orientation, not as a recommendation:
- SAP Ariba / SAP Business Network: With roughly 4 million buyers and suppliers in over 190 countries and an annual transaction volume of around 3 trillion USD, SAP Business Network is the largest network worldwide (SAP Business Network). Both OCI and cXML are supported.
- Coupa: With more than 10 million buyers and suppliers and 8 trillion USD cumulative spend, Coupa is the dominant cXML platform in the enterprise segment (Coupa).
- JAGGAER:27 percent of DACH top companies use JAGGAER (JAGGAER) - especially strong in public procurement and corporations with complex supplier networks.
- Onventis (Stuttgart): With more than 1,000 procurement organizations, 1.2 million users and around 20 billion USD annual transaction volume, Onventis is the leading cloud procurement platform in German SMBs.
- Unite / Mercateo: The marketplace connects 1.5 million commercial customers with 16,000 manufacturers and 33 million articles - as an aggregator it is often the simplest entry option for suppliers without ERP punchout experience.
- Oracle Procurement Cloud: Standardized cXML stack, in DACH mostly present in subsidiaries of US corporations.
For new projects we recommend starting with the system your largest buyer already uses - this usually yields the most economically valuable quick wins. If you want to open up new customer segments deliberately, starting with Onventis or SAP Business Network is a safer bet, as those platforms reach the broadest DACH mid-market.
Implementation in 5 steps
A punchout project cannot be done in a week, but it is also not a mammoth project. The following step-by-step outline shows how we typically approach it at XICTRON:
- Requirements workshop (1-2 weeks): Which buyers are relevant? Which platforms do they use? Which articles should appear in the catalog? Which prices and volume tiers apply? Which additional fields (cost centers, project numbers) does the cart return have to carry? This phase is decisive - mistakes here propagate through the entire project.
- Prototype and format decision (1 week): OCI or cXML? Which version? Which authentication? We build a minimal prototype with the largest target buyer system to validate data structures - before any production code is written.
- Shopware integration and B2B Components configuration (3-5 weeks): Create customer groups, set up pricing rules, assign assortments, build session handling and SSO flow. The B2B Components in Shopware 6 form the foundation - custom development only happens in targeted places for punchout-specific endpoints.
- Bidirectional ERP and OrderRequest (3-6 weeks): The connection to SAP Business One, Microsoft Dynamics or another ERP. Prices, stock and customer data from ERP, orders back into ERP - depending on the system this is one to several interfaces.
- End-to-end testing, pilot and rollout (2-4 weeks): Real test orders with a pilot buyer, catching edge cases (such as session timeout, duplicate cart returns), then gradual rollout to additional buyers.
For a first punchout connection to a platform you should plan 3 to 4 months - from requirements to pilot go-live. Every additional platform can then be integrated significantly faster, because the core architecture is in place. A project that needs to cover several platforms in parallel realistically takes 5 to 7 months.
How XICTRON implements your B2B integration
As B2B e-commerce specialists based in Lower Saxony, we guide mid-market companies and corporate subsidiaries through the entire path from concept to a production-grade punchout flow. We bring both the technical expertise for Shopware integrations - including SAP, Microsoft Dynamics and other ERP systems - and an understanding of the procurement processes on the buyer side.
Our approach stays pragmatic: we do not build punchout solutions for a showcase, but for measurable efficiency gains. In our experience, well-planned punchout projects typically pay back within 12 to 18 months - especially when the B2B pricing strategies, ERP processes and the product configurator are tuned properly together. Further strategic aspects - for example how you work on B2B digitalization in sales - are covered in our adjacent articles.
This article draws on data from: Capital One Shopping Research (B2B E-Commerce Market), BEVH Impact Studie 2025, Trade.gov DE B2B Report, Mordor Intelligence, BME/Onventis Procurement Monitor 2024, Bitkom digitalization studies 2024/25, Greenwing Technology, TradeCentric, Digital Commerce 360, Procurify, APQC Open Standards Benchmarking, Ardent Partners State of ePayables 2025, Hackett Group Procurement Benchmark, Ivalua/Sievo (Maverick Spend), Aberdeen Group, Fairmarkit, Conexiom, Sana B2B Buyer Report 2025, McKinsey B2B Pulse Survey 2024, Forrester Predictions 2025, Gartner Future of Sales, Shopify Enterprise 2025, Apps Run The World, Global Market Insights, SAP Business Network, Coupa, JAGGAER, Alumio, Silk Commerce, Shopware/PR Newswire. Concrete figures may vary by industry, region and time of measurement.
Frequently asked questions
No. While the biggest deployments do happen in large corporations, mid-market companies increasingly use e-procurement platforms such as Onventis or Unite. If you have several buyers who want to order out of an ERP, a punchout catalog typically pays off starting from an annual order volume in the mid six-figure range. Even the customer lifetime value of your B2B customers benefits measurably from a smooth connection.
In practice we usually support both formats in parallel, because target platforms differ. OCI dominates in DACH with SAP-driven buyers, cXML is common on Coupa, Oracle and with US-oriented customers. The actual Shopware logic (session, customer pricing, cart return endpoint) is reusable - only the protocol layer differs.
The B2B Components in Shopware 6.8 deliver the foundation for customer groups, quick order, volume pricing and approvals - exactly the functions a punchout buyer needs in the shop. The punchout flow builds on top and extends the shop with setup request handling and cart return endpoints.
In the setup request: buyer identification, a HOOK_URL (where the cart should return) and optionally language/currency. In the cart return: all cart line items with quantity, price, SKU, short description and optional metadata such as cost centers, project numbers or delivery addresses. cXML supports significantly more structured additional fields than OCI.
That is normal: not every cart gets approved. Your shop should log punchout sessions properly, track abandoned flows as an analytics signal and stabilize prices for a limited time - so the final OrderRequest does not end up on outdated values. We typically implement a cart lifetime of 24 to 72 hours.
After the initial go-live the ongoing effort is manageable. Typical activities are monthly monitoring (session rates, error rates, cart return success) and occasional adjustments when the platform side changes or new buyers are added. Critical areas are price accuracy, assortment maintenance and DATEV accounting processes - we therefore recommend a fixed maintenance contract for production punchout installations.
In 2026, punchout is no longer a nice-to-have - it is the entry ticket into professional B2B procurement processes. Anyone who connects their Shopware shop cleanly to OCI and cXML makes themselves strategically indispensable for their buyers' procurement teams - and at the same time gains a measurable efficiency dividend. We support you through every step, from the first setup request to the productive OrderRequest.