SAP promises to open out data flows in both directions

SAP added native Apache Iceberg support to Business Data Cloud and expanded its Databricks partnership, enabling governed, zero-copy data sharing across platforms including Snowflake, Google, AWS and Microsoft. The changes also aim to connect legacy ECC data and relational databases to cloud analytics and AI tools without moving data, easing ERP-upgrade constraints and reducing vendor lock-in. Analysts see the greater openness as beneficial for customers, though control of the enterprise context layer remains contested and implementation work lies ahead.

Analysis

The strategic value is less “open data” than a possible decoupling of AI/analytics adoption from ERP migration. If ECC customers can build useful workloads on SAP business semantics before moving to S/4HANA, SAP may protect account relevance and create a lower-friction path to BDC consumption. That could reduce upgrade urgency while improving long-run retention: a near-term headwind to migration-led monetization, but a potential extension of SAP’s customer lifecycle.

The countervailing effect is that Iceberg and zero-copy access weaken proprietary data-format lock-in. Databricks, Snowflake and hyperscalers can compete for the analytics runtime and, more importantly, the business-context layer; SAP’s moat is strongest only where its semantics materially improve model quality or agent actions. Databricks’ expanded role validates the ecosystem but also makes it a potential value-capture rival. Dremio’s technology may improve interoperability, but the article provides no evidence yet of customer adoption, paid workload growth, or economics.

Immediate market impact should be limited absent monetization proof. Over 1–3 months, watch BDC customer/workload disclosures, ECC-to-BDC use cases, and whether Joule–Genie integration produces governed actions rather than demos. Over 6–18 months, success could support cloud and AI attach without forcing ERP conversion; failure leaves SAP funding integration complexity while partners capture usage. The contrarian risk: openness is being read as a moat enhancer, when it may instead make the data layer more substitutable and shift bargaining power to customers.

Key Decisions for Investors

• Treat SAP as a conditional, medium-term positive rather than buying the announcement alone. Consider adding only on weakness if subsequent reporting shows BDC workload or cloud-revenue traction; the thesis is falsified if adoption remains mostly migration-independent pilots without measurable commercial contribution.
• Track whether ECC customers actually deploy production analytics/AI workloads through BDC, and whether those workloads persist through S/4HANA migrations. Verify customer counts, paid consumption, and renewal/attach disclosures; do not infer revenue from the conference examples.
• Monitor Databricks, Snowflake and hyperscaler positioning as both distribution partners and substitutes. If customers increasingly choose external context layers and runtimes while SAP supplies only governed source data, reduce the expected SAP AI/data-platform premium.
• No high-conviction pair trade is warranted from this announcement alone: there is no quantified impact, valuation context, or evidence of competitor displacement. Reassess after the next SAP results or product-adoption disclosures.