Over 430,000 enterprises worldwide rely on SAP ERP systems to power their operations, including 98 of the world’s 100 largest companies. SAP systems serve as the foundation for many businesses, managing core processes like finance, inventory management, customer relationship management, and human capital management. Data from all of these systems is invaluable, offering the potential for a comprehensive view of an organization’s operations, decision support, and innovation of all kinds.

While SAP products are highly customizable, the SAP ecosystem is closed by design, a classic example of how vendors can lock-in their customers. Customers often struggle to access and utilize their own data on third-party platforms, which complicates analytics and hinders innovation. Instead, SAP pushes customers to use their underdeveloped, proprietary data platform.

Our team has spoken with several global enterprises who want to integrate their SAP data into modern data lakes or warehouses. While there are multiple paths, such as using a third party platform for analytics, storage, and more, customers feel forced to choose between:

• Paying a premium for an enterprise SAP HANA database license to access their data
• Using Datasphere as a costly ELT tool, with extra charges for outbound data
• Building a federated architecture in Datasphere that struggles with scale

Shrinking customer choice

Over the course of successive product releases, SAP has taken clear steps attempting to narrow how customers extract data, guiding users toward SAP-certified products and away from generic ELT/ETL solutions. This approach has continued with its current flagship product, SAP S/4HANA, first launched in 2015, and has since intensified.

In 2022, SAP note 2971304 raised concerns around log-based replication. In a 2023 blog, SAP stated outright that it would no longer certify non-SAP data integration vendors, which effectively limited the tools customers believed they could use to access their data. Note 3255746, released in July 2024, sought to prohibit the popular approach to use ODP (operational data provisioning) via RFC (remote function calls). These mid-stream updates signal a growing insistence on SAP’s proprietary solutions over tried-and-true third-party options.

On ECC and S/4 HANA, SAP allows customers to directly access their database to extract SAP data, but only if customers have purchased an enterprise database license. This license is often priced several times more than the runtime license, introducing unexpected fees that can derail budgets and catch customers off guard.

SAP’s product upgrade paths increasingly push customers toward RISE with SAP, their cloud-based managed service. RISE is sold most commonly with a restricted runtime license that does not allow direct database access; instead, customers must use SAP-defined interfaces. Direct access and support for third-party extraction tools therefore are not included by default and require the purchase of an enterprise license. This limits customers’ ability to freely access, extract, and manage their own data and continues to illustrate SAP’s long-standing strategy of keeping customer data tied to its platform.

 While this approach may have secured customer loyalty in the 1990s, it is outdated. Modern companies, such as Salesforce, have demonstrated that providing customers with more flexibility over their data actually strengthens customer loyalty and platform value. SAP should not be afraid of openness and interoperability; customers running SAP for ERP or other critical systems are unlikely to leave because of the inherent strength and importance of SAP's services. If SAP adopted more open practices, customers would be able to derive greater value from their SAP data in modern data platforms, which would, in turn, make SAP itself more valuable.

SAP Datasphere

To address customers’ growing need to integrate and analyze data across systems, SAP released Datasphere (formerly SAP Data Warehouse Cloud) as the central hub for SAP and non-SAP data integration. SAP funnels customers toward Datasphere as the only viable data platform, especially when it comes to accessing newer features like AI and advanced modeling.

SAP is positioning Datasphere as the future of data warehousing and analytics while gradually phasing out legacy tools like SAP Business Warehouse. Similarly, support for ECC is set to end in 2027, with optional extensions until 2030. To extend ECC maintenance through 2033, customers must adopt SAP RISE, effectively requiring a migration and limiting options for customers who prefer to remain on-premise or use hybrid architectures. These moves reflect SAP’s broader strategy to consolidate its customers within its own cloud ecosystem.

Whether it’s analytics, dashboarding, or GenAI, organizations need to centralize both SAP and non-SAP data. Exporting data from Datasphere to non-SAP targets, however, often incurs additional costs. Combined with the sunsetting of tools like SAP Business Warehouse and the migration push to SAP RISE, customers are left with fewer choices for the technical implementation of their data strategy and steeper price tags.

Datasphere compares poorly with modern data platforms

Compared with third-party data lakes and data warehouses such as Snowflake, Databricks, Google BigQuery, and Redshift, Datasphere lacks cost-effectiveness, flexibility, interoperability, and performance, as well as lacks certain analytical capabilities and handling of unstructured dataAcross multiple customer conversations, several recurring themes have emerged.

Exorbitant costs

Datasphere features a fragmented pricing model, including piecemeal, unpredictable charges for compute, storage, catalog, data lake access, and other features.

Poor interoperability

Upstream, Datasphere prioritizes SAP-native integrations and offers a very limited number of inbound and outbound connectors, which customers must pay for separately.

Downstream, Datasphere doesn’t operate well with third-party tools and platforms. Datasphere’s use of CDS views and HANA-specific modeling prevents interoperability with open-source transformation and modeling tools like dbt. Schema changes are brittle, requiring manual edits to SAP views, even for simple tasks such as adding a column. Only DDL or limited view definitions are valid, and there are gaps in SQL function compatibility. 

The picture is no better for data visualization and business intelligence. One customer reported having to build multiple layers of views to get simple dashboards to work in Tableau because Tableau cannot directly query tables in Datasphere because of SAP’s specific modeling. 

Even with zero-copy, integrating huge data sets with low latency requirements can be challenging. 

Architectural limitations

Datasphere’s compute and storage resources are included together in what SAP calls the “core application,” which contributes to capacity unit usage. While SAP does not specify whether these resources can be scaled independently, the bundled pricing model implies limited elasticity. This can lead to inefficiencies in dynamic workloads where compute or storage demands vary independently.

In addition, customers have reported performance challenges in SAP Datasphere related to memory allocation, particularly when executing complex transformations. In some cases, insufficient or inconsistent memory handling has led teams to offload more demanding workloads to other platforms. As one customer put it, they “still lean on Snowflake because it handles memory allocation more reliably.”

Modern data platforms are built to meet enterprise data needs

For all the aforementioned reasons, many organizations still prefer to centralize SAP data in data lakes and warehouses like Snowflake, Google BigQuery, Databricks, or Redshift, as these platforms offer: 

• Cost effectiveness and scalability. Transparent pricing based on usage with predictable compute/storage billing and fewer unexpected surcharges.
• The ability to handle large data sets from diverse workloads in real-time and support rapid schema changes
• Interoperability across cloud providers, SaaS applications, major BI platforms, and more. 

All critical capabilities for AI and advanced analytics.

Recent SAP Accouncements 

The February 2025 SAP Databricks announcement, integrated within the SAP Business Data Cloud (BDC), is currently limited to SAP RISE Private customers. BDC also introduces a subscription-based pricing model that is separate from SAP’s Business Technology Platform Enterprise Agreement (BTPEA). Although Databricks is included within the BDC licensing structure, it represents an additional investment on top of existing SAP Datasphere expenditures. The integration can also introduce architectural complexity, organizations must navigate two distinct ecosystems, potentially complicating data governance and interoperability.  

Where does this leave SAP customers?

SAP cites system integrity and specialized support to justify these constraints. But if the goal is truly performance and customer success, why impose so many barriers to data freedom or push customers into a closed ecosystem that limits integration with best-in-class tools?

The reality is: customers want — and deserve — choice. Data teams need the freedom to build the best data architecture for their goals, whether that includes SAP or not. That’s where Fivetran comes in.

Fivetran offers a modern approach to SAP data integration, enabling customers to centralize both their SAP and non-SAP data in cloud data platforms with minimal engineering effort. We help enterprises avoid vendor lock-in, lower their total cost of ownership, and accelerate time to insight.

In the next part of this series, we’ll explore the practical paths forward for SAP customers looking to modernize their data stack, navigate restrictive licensing, and unlock true agility on their terms.

[CTA_MODULE]