Why Product Data Sets the Ceiling for SAP AI Investments

SAP is giving AI assistants and agents a larger role in product recommendations, quotations, and customer service. For manufacturers, brands, and retailers, that raises a practical question: what information will those agents use?

“AI doesn’t create trustworthy product knowledge. It consumes it,” Dominic Citino, senior vice president of corporate development at Inriver, told SAPinsider in a recent podcast. SAP manages the operational product record, but agents also need product data from outside the core to describe, configure, and sell products.

“The quality of the data sets the intelligence ceiling” for every agent using it, Citino said. If product information is incomplete, inconsistent, or siloed, “the AI layer doesn’t fix that. It amplifies it — at scale and speed.”

That changes the question for SAP customers. “The traditional question was, ‘Should I buy a PIM?’” Citino said. “The question SAP customers are now asking is, ‘How do I ensure my SAP AI agents have trusted product knowledge?’” Inriver’s answer is a governed product-data layer connected to SAP but maintained outside the ERP core.

SAP Keeps the Record; Inriver Adds the Context

Inriver’s architecture separates the operational product record managed in SAP from the richer product data used across commerce, service, search, and AI. Its SAP-certified BTP ERP Cloud Connector links the two, allowing SAP to remain authoritative for operational data while Inriver governs enrichment outside the SAP S/4HANA core.

“SAP owns the operational product record. Inriver owns the commerce-ready product data,” Citino said. That helps keep SAP’s operational records aligned with the product data used by commerce and AI as products change.

“Because the integration is bidirectional, these two systems stay synchronized without custom code,” Citino said. Inriver positions that model as an alternative to embedding product-content workflows in ERP or building separate point-to-point connections for each downstream application. Product data managed in Inriver can then support SAP Commerce Cloud, external sales channels, and AI applications.

The Connector currently applies to cloud ERP environments, including SAP S/4HANA Cloud and RISE with SAP, rather than fully on-premises SAP ECC systems. For ECC customers, Citino said, “This isn’t a ‘connect today’ story yet — it’s a ‘build toward this’ story.”

Four SAP Scenarios Put Product Governance Into Practice

Inriver sees four points where SAP customers may need to reconsider how they manage product data. Each starts with a different technology or business requirement, but Citino said they “all lead to the same strategic outcome”: a product intelligence layer that supports current commerce operations and future AI use.

The first is an SAP S/4HANA migration. “An S/4HANA migration is about much more than modernizing ERP,” Citino said. Inriver advises customers to design the operational and product-data layers together, rather than complete the ERP migration and address product content later. SAP can then manage the operational record while Inriver prepares the data used by commerce and other customer-facing systems.

For organizations already using SAP Commerce Cloud, the immediate issue is the quality of the product data feeding the platform. “Every commerce platform is ultimately dependent on the quality of the product data it receives,” Citino said. Inriver’s platform is where data is enriched and governed before it reaches Commerce Cloud, helping keep product information complete and consistent across the customer experience.

A third scenario arises when organizations outgrow SAP Product Catalog Management or another catalog tool. “PCM works well for relatively straightforward product catalogs,” Citino said. But additional markets, sales channels, product variants, and governance requirements make that information harder to manage. A dedicated product-data platform can provide the workflows and controls needed as catalog complexity increases.

“The fourth scenario reflects where the market is headed,” Citino said. As SAP develops cloud-native, AI-enabled commerce, organizations still face the same question: “Where does the AI get trusted product knowledge?” Those agents depend on product data that extends beyond SAP’s operational record. Citino argues that establishing that foundation now prepares customers for AI taking on more customer-facing work.

Product Data Becomes a Cross-Functional Responsibility

“This is no longer just an IT conversation,” Citino said. Finance and executive leaders are asking “how do we maximize the return on our SAP and AI investments while reducing operational risk?” Product data affects revenue, compliance, customer experience, and whether automated processes work as intended.

Product, marketing, and commerce teams approach the issue from another direction. They need product information that is accurate, complete, and consistent wherever customers engage, Citino said. AI leaders face a related question: “Is the organization ready for what’s next?” For both groups, product-data readiness determines whether new customer experiences and AI capabilities can work as intended.

“Start with an honest assessment of your product data,” Citino said. Customers should ask where it lives, how complete it is, and how many “versions of the truth” exist across the organization. They should also identify which customer experiences and AI capabilities depend on that information. Citino points out that that those questions can reveal where ownership and governance are weak, and where a more consistent product-data foundation could create measurable business value.

The Connector is only part of that foundation. “Technology platforms will evolve. Commerce channels will evolve. AI capabilities will evolve,” Citino said. SAP customers should not have to rebuild their product-data strategy each time. Inriver’s argument is that a governed layer outside the ERP core can support current commerce operations while giving future agents product information they can trust.

What This Means for SAPinsiders

Product data sets AI’s ceiling. SAP customers need more than operational records; they need governed product context that agents can use reliably. Inriver positions its connector as the boundary between SAP’s core and that enriched data.

SAP S/4HANA migration provides a decision point. Designing the operational and product-data layers together prevents governance from becoming a post-migration repair. That allows SAP customers to establish trusted data as part of the transformation, giving commerce and AI teams a stronger foundation when the new SAP environment goes live.

Product data is becoming a shared investment. Because data ultimately affects revenue, compliance, customer experience, and AI performance, ownership cannot remain with IT alone. A governed product-data foundation can support SAP investments as sales channels and AI capabilities evolve with the business.