Florian Bigelmaier
Analyst Data & Analytics
Florian Bigelmaier is an analyst with a focus on data management and data intelligence.
Build data platforms that last
Independent guidance for architecture, vendor selection, and modernization. Based on 10,000+ user voices per year and 500+ vendor and product analyses
25+ years of independent market research · 10,000+ user voices per year · 25 analysts
The data foundation for AI and analytics
A data platform stores, integrates, and delivers data, making it usable for analytics, BI, and AI models. Choosing the wrong architecture in 2026 means building on shaky ground for the next three to five years. Several forces are now converging at once:
- SAP BW is reaching end of life.
- Hyperscaler stacks are maturing at different speeds.
- Data sovereignty is reshaping architectures.
- AI workloads are pushing classic BI topologies past their limits.
- Unified access to distributed data is now a baseline expectation.
For platform owners, CIOs, and CDOs, this means seven- to eight-figure investments are at stake, with a direct impact on architecture, operating model, and AI roadmap.
For 25 years, BARC has evaluated data platforms independently, with no implementation work, no reseller deals, and no vendor commissions on the line. On this page, you’ll find what you need to make that call: our data platform consulting offerings, the situations that typically bring platform owners to us, the criteria that separate a sustainable platform choice from a costly one, and the analysts who will work alongside you on your project.
Platform decisions in 2026 are a strategic question, not a technical one
Whether a company gets real value from its data and AI comes down to how it selects and evolves its data platform. Six findings from our current research illustrate this:
- The share of users reporting no significant platform problems has fallen from 40% to 26% in two years.
- Microsoft Fabric (26%), Databricks (21%), and Snowflake (19%) dominate current evaluations. Specialists such as Dremio and Exasol offer clear alternatives for performance and sovereignty requirements.
- 51% of companies consider data sovereignty very important, up from 42% the previous year. 76% expect its importance to keep rising.
- The hyperscaler push for a “European Sovereign Cloud” is missing its target: only 10% of companies rely on it, fewer than the 13% slowing their cloud migration or the 16% pursuing full repatriation. The market preference is clear, and it is not hyperscalers wearing a sovereignty veneer.
- Data and AI in core processes drive 62% of all sovereignty initiatives.
- 45% of companies name data quality as the biggest obstacle to AI, more than twice as many as in 2024.
The gap between vendor promises and actual user experience is growing. Today’s platforms most often fall short in three areas: cost transparency, governance enforcement, and AI readiness.
What users face today, and where the journey leads
| What users experience today | Target state |
|---|---|
| "Our Snowflake bill has doubled in twelve months without the workloads growing accordingly." | Full cost transparency and a negotiable pricing model |
| "Self-service works for power users, the rest of the business departments are left out." | An architecture that measurably enables business teams to self-serve |
| "We do not know whether our current platform will still fit our AI strategy in two years." | A reliable target architecture for GenAI, agents, and unstructured data |
| "For every new data source, we build the connector ourselves." | An integration concept that covers SAP and non-SAP sources as well as structured and unstructured data |
| "Our auditors ask about data sovereignty, but we cannot demonstrate it at the architecture level." | A documented sovereignty architecture without a rebuild when switching platforms |
| "Governance exists on paper, but nobody enforces it operationally." | Enforced governance that is visible in the platform and does not gather dust as a PDF |
| "We operate three overlapping platforms, and none replaces the other." | A consolidated multi-platform architecture with clearly defined areas of responsibility |
| "Our SAP BW is being phased out in 2027, we are migrating to S/4HANA in parallel, and a robust successor architecture is missing." | A decision-ready evaluation of SAP BDC, Microsoft hybrid architecture, and third-party options |
| "IT wants to stick with BW4 and points to SAC performance, without any reliable comparative figures." | Performance and TCO comparisons between BW4, Datasphere, and hybrid architectures instead of vendor arguments |
| "We have two separate worlds: an SAP DWH for BI and Databricks for AI, with no clear data ownership in between." | An integrated architecture with a data product and ownership model that brings BI and AI workloads together |
Typical triggers for a data platform inquiry at BARC
The inquiries that platform owners bring to BARC in 2026 can be traced back to three recurring triggers. They often overlap, particularly the first two.
Trigger 1: SAP BW is reaching end of life. Maintenance runs out in 2027 (BW 7.x) or 2030 (BW/4HANA with extended maintenance), an S/4HANA migration is already underway, and the core question is still open: whether to stay with SAP at all, and whether to run in the cloud or on-premises. IT often holds on to BW4 because many stakeholders still see the performance concerns around SAC as unresolved. On top of that come tool consolidations (Qlik, Power BI, SAC, AFO, Corporate Planner, Longview) and pressure to put planning and reporting on a single data foundation.
Trigger 2: SAP-Microsoft hybrid architecture as the target state, evaluation still open. SAP Business Data Cloud (Datasphere) plus Microsoft Fabric or Databricks is the most common architecture companies bring to BARC in 2026. They want to know how the options (ODP/OData straight to Azure versus processing through Datasphere) compare on cost, licensing, effort, and skill requirements, and how to decide the front-end strategy between Power BI and SAC in practice.
The operational questions matter just as much: How do I sync data or data products between SAP and other systems? How do I establish a unified access layer? Where do I store which data? And which access and exchange options are allowed, and which are not?
Trigger 3: Consolidation of separate BI and big data worlds. Architectures that have grown over the years often consist of distributed platforms, for example an SAP DWH alongside a Databricks lakehouse on AWS or Azure. The result is redundant data storage, shadow IT through Power BI, and no clear data ownership. BARC is then tasked with designing an integrated target architecture, defining data product and ownership models, and laying out a realistic modernization path with quick wins.
Three constants run through all triggers: the decision has to be made within 6 to 18 months, the investment sits in the seven- to eight-figure range, and the stakeholders range from the CIO to controlling and the data governance board.
Who we talk to for a data platform project
In data platform selection and architecture projects, BARC typically works alongside the following roles:
- Platform and architecture responsibility: Enterprise Architect, Business Architect Data Platform, Head of BI & Analytics, Head of BI and Big Data Platform, Director Big Data Solutions, Lead AI Manager.
- Data governance and management: Head of Data Management, Head Data & Digital Experience, Manager IT Governance and Transformation, Lead AI/ML Data Scientist, Data Management Expert.
- Executive sponsorship: CIO, CDO, CFO, Vice President Digitalization.
- SAP-specific roles around the BW end-of-life: SAP Program Manager, Transformation Manager, Manager ERP Applications Business Intelligence, Head of Chapter Reporting & BI Solutions, Head of Group Controlling, Head of Finance and Controlling.
Industries and company sizes: midmarket companies and large enterprises across all industries, with a focus on manufacturing, retail, financial services, and utilities.
How BARC helps you choose the right platform
What matters in 2026 is the platform topology that will carry your data and AI strategy for the next five years. The individual vendor is secondary. The market is moving in three directions at once: consolidation toward the three hyperscaler-adjacent platforms, specialists with clear advantages for specific workloads, and growing sovereignty pressure that puts hybrid and European stacks back on the agenda.
BARC has evaluated platforms with the same methodology for 25 years: user surveys, vendor assessment, architecture review, and economic analysis. We earn no implementation revenue, run no reseller business, and take no vendor commissions. That independence is exactly why our recommendations hold up at the negotiating table.
In platform decisions worth six to eight figures, BARC delivers three contributions that set it apart from internal evaluation, Big 4 consulting, implementation partners, and specialist consultancies:
- Facts, not opinions. Market data from 10,000+ user surveys a year, plus continuous evaluation of the vendor pipeline that matters to your decision.
- Less time spent, less risk taken. A structured selection method with proven requirement catalogs, shortlist logic, and economic models.
- You negotiate on equal footing. Negotiation benchmarks for pricing and contract models.
Your options with BARC
Decision criteria for a sustainable data platform in 2026
A sustainable data platform decision in 2026 answers seven questions with a clear answer:
- Workloads. Which use case classes will run on the platform over the next three years: BI, analytics, ML, GenAI, agents, real-time?
- Data topology. Where is the data created, where may it reside, where does it have to be processed? Hybrid, single-cloud, or European sovereignty zone?
- AI architecture. Vector and graph stores, AI governance, semantic layer, MCP connectivity: integrated or separate?
- Governance and sovereignty. Which regulatory requirements and policies (data and AI governance) are active today, which are coming in the next 24 months (EU AI Act, NIS2, data protection), and how are they enforced operationally in the platform?
- Economics. Which pricing model (compute-based, storage-based, user-based) fits your usage and your load curve?
- Operating model. Central platform organization, domain ownership, or a hybrid model? Who is responsible for self-service?
- Integration and semantic layer. How are structured and unstructured data (documents, images, logs, SAP and non-SAP sources) integrated, and through which semantic layer and which cataloging concept are they made consistently available for BI, analytics, and AI?
We recommend answering each of these seven questions with an explicit architectural statement before the vendor shortlist is created. Without these seven answers, software selection becomes a feature list.
Frequently asked questions about data platforms
Data platforms are software systems for managing the integration and provision of data and its metadata. They combine functions for data storage, integration, processing, data access (for example query engines, APIs, federation/virtualization), a semantic layer, governance, analytics, and AI in a unified architecture, with the goal of making data available consistently, securely, semantically aligned, and enterprise-wide for analytical, operational, and AI use cases.
Modern data platforms support architectural approaches such as lakehouse, data fabric, and data mesh, enable the creation of data products, and integrate data intelligence components such as data catalog, governance, and observability.
A data lake stores raw data in open formats without a fixed schema. A data warehouse stores structured, modeled data for analytics. The lakehouse separates storage and compute and uses open table formats such as Apache Iceberg or Delta Lake. Data fabric is an architectural pattern that logically brings together distributed data sources through metadata and integration. It provides the architectural basis in which business departments own data products as independent units of responsibility.
The right platform depends on workloads, data topology, AI requirements, and governance profile. Microsoft Fabric is strong for Microsoft-centric organizations and self-service. Databricks leads for ML and lakehouse workloads. Snowflake offers operational maturity and easy scaling. SAP Business Data Cloud is relevant for SAP-centric companies. Dremio and Exasol are strong alternatives for specific performance and sovereignty requirements. A reliable decision follows from a structured selection rather than from a vendor pitch.
A lakehouse is worthwhile as soon as unstructured data, ML workloads, or a growing number of AI use cases place demands on the architecture. For purely structured BI workloads, the data warehouse often remains the simpler and cheaper choice.
Most companies ultimately choose a hybrid architecture. A cloud data warehouse carries the governance-intensive workloads, and a lakehouse layer processes ML, streaming, and unstructured data. Details on use cases and adoption paths are in the BARC Lakehouse Cookbook: Practical Business Use Cases And Adoption Pathways.
Three levers:
- Clarify the workload profile before choosing the platform: what load do we expect, which pricing model fits it?
- Enforce governance: through compute quotas, limits, and automatic pausing of unused workloads.
- Negotiate pricing using volume benchmarks from ongoing vendor coverage data.
For each of the three levers, BARC provides user data and negotiation benchmarks from the ongoing vendor coverage.
AI readiness rests on four pillars: integrated vector and graph stores for retrieval-augmented generation workloads, a semantic layer for consistent business logic, AI governance functions for model and prompt lineage, and end-to-end data quality that protects generative systems from hallucinations. In the BARC Data, BI and Analytics Trend Monitor 2026, 45% of companies name data quality as the biggest obstacle to AI, more than twice as many as in 2024.
No. BARC does not implement, integrate, resell, or take vendor commissions. That separation is exactly what keeps our recommendation reliable at the negotiating table. For implementation, we recommend independent partners from our ongoing BARC Data & Analytics Service Provider Survey.
A workshop for a strategic situational assessment takes one day. A full software selection with shortlist, demo evaluation, and contract recommendation usually takes 6 to 12 weeks, depending on the scope and vendor.
BARC receives no commissions or license shares from software vendors. The vendor evaluations are based on user surveys with more than 10,000 voices per year and on analyst reviews. We disclose sponsorship of BARC studies openly, and it does not influence the evaluation methodology.
BARC Consulting covers data strategy, data management, AI governance, integrated planning & analytics, corporate performance management, and ESG. An overview of all consulting offerings can be found here.
No. BARC structures the selection so that we include and evaluate suitable implementation partners directly in the candidate pool. You leave the process with two recommendations: software platform and implementation partner. Verified partner ratings per platform are provided by the ongoing BARC Data & Analytics Service Provider Survey.
The concern is valid. BARC addresses it in two ways. Our senior analysts have worked for 10 to 25 years with exactly the platforms you are evaluating today. From 500+ projects they have supported, they know the typical implementation pitfalls. Our results provide an implementation roadmap, the required skill mix, pricing negotiation benchmarks, and concrete next steps for the first 90 days after the contract is signed.
Then the Data Platform Guidance Package is your pragmatic entry point. One workshop day works through the seven core questions (workloads, topology, AI architecture, governance, economics, operating model, integration and semantic layer) and delivers a calibrated situational assessment. With this result, you can reliably justify the investment for a full selection or a health check internally.
Who trusts BARC
Read our customer success stories to see how we help our clients achieve their goals.
The BARC data culture workshop was very insightful, particularly the original industry survey results, which offered valuable benchmarks. It helped our participants gain a better understanding of how to measure and improve data culture, drive organizational changes, and become better data leaders. A great opportunity for organizations looking to foster a sustainable data culture and elevate their decision-making processes.
Dr. Wen (Jenny) Shi
Lead Data Scientist, zeroG
We had the pleasure to have Jacqueline as our keynote speaker for our internal Data Science Conference uniting the international community of data enthusiasts at Boehringer Ingelheim. Jacqueline highlighted the potentials and recommendations to implement Data Mesh in corporate environments. Reflecting on the high importance of innovation for the pharmaceutical industry and Boehringer Ingelheim in particular, I can only emphasize the excellent quality of her presentation and in-depth knowledge reflecting the expertise BARC is commonly known for.
Heiko Merten
Head of Data Science Academy & Product Teams, Boehringer Ingelheim
BARC research reports help focus on trends and provide supporting data and facts! Especially if you don’t know yet whether you should or even have to focus on a topic, BARC offers excellent advice from specialists!
Barbara Engerer
Global Digital Services, Head of Data and Analytics GDS-DA, BSH Hausgeräte GmbH
I have been using the Trend Monitor for years to show my students what is happening in the business world with data, BI, and analytics. I also use the data directly, using the top trends to discuss key issues and challenges, as well as the background to the buzzwords.
Prof. (FH) DI Dr. Heidi Weber
Lecturer, Vorarlberg University of Applied Sciences
10,000+ user reviews per year · 500+ vendor analyses per year · 25 years of methodology
These companies trust us
Your experts for data platforms
Timm Grosser
Senior Analyst Data & Analytics
Timm Grosser is a Senior Analyst Data & Analytics with a focus on data management.
Kevin Petrie
VP of Research at BARC US
Thomas Zeutschler
Senior Analyst Data & AI
As an analyst and technology expert, Thomas supports his clients in adapting modern data & analytics concepts and technologies and in developing data-driven processes and business models.
Featured Insights
Don‘t miss out!
Join over 25,775 data & analytics professionals and get the latest product insights, research, surveys and more!
Events
Deciding on your data platform in the next twelve months?
A 30-minute BARC briefing clarifies whether a Guidance Package, a full software selection, or a health check fits your situation. You leave with a clear next step, regardless of whether we continue working together afterward.