How BMW Group Uses Master Data Service for Non-Tset Applications

At our first edition of the Tset Summit 2025, together with BMW Group, we discussed how its organization approaches cost engineering across an unusually broad scope. Their presentation revealed a strategic approach to managing cost data that extends far beyond typical automotive applications.

Cost Engineering Across the Enterprise 

Most manufacturers apply cost engineering primarily to vehicle components, but BMW Group takes a different approach. Their cost engineering function spans direct materials, indirect goods, services, machinery, facilities, and even marketing events. This breadth creates unique challenges that require different methodologies and tools for each category. 

The scope reflects a fundamental belief: if purchasing touches it, cost engineering should analyze it. This includes everything from aluminum body components to plant buildings, from logistics services to software licenses for employees. Each category demands its own calculation methodology, yet all must draw from consistent foundational data.

The Challenge of Multiple Software Solutions 

Applying cost engineering across such diverse categories creates a fundamental challenge: no single tool can effectively handle all use cases. For example, service costing requires fundamentally different structures than component costing; and concept phase estimations need different approaches than detailed bottom-up calculations for supplier negotiations. 

Heavily tailoring standard software to specific company demands creates maintenance challenges that become unsustainable for software suppliers. This realization led to a "clean core" policy for standard software, while keeping modifications minimal while building flexibility through other means. 

The solution involved embracing multiple tools while establishing a centralized data foundation. This required thinking differently about system architecture.

API-First Architecture as the Foundation 

BMW Group's approach centers on an API-first architecture for their master data foundation. This technical decision enables something strategically important: the ability to build custom applications on top of standardized data without modifying core systems. 

All functionalities remain accessible from non-Tset applications throughout the software development process, enabling seamless data flow and transformation across different systems. 

Multiple scenarios become possible with this approach. Teams can develop cloud-based applications tailored to specific BMW Group workflows while pulling from centralized cost engineering data. Semi-automated tools can assess supplier offers by fetching calculation premises from the central repository. A data management cockpit serves users across functions with live monitoring and analytics capabilities. 

The key advantage is the fact that end-to-end data flows without requiring everything to run in a single system. Different clouds, different tools, same underlying data foundation.

Opening Data Across Functions 

Perhaps the most significant shift involves who can access cost engineering data. Historically, cost engineering departments often kept their data closely held. BMW Group is moving in the opposite direction. 

The organization has committed to opening cost engineering data to other divisions - a departure from past practices where departments maintained isolated datasets with little willingness to share information across functions. 

This represents both a technical and cultural change, starting with location factors like wages that multiple departments need, then gradually expanding data accessibility over time. The approach acknowledges that some datasets carry more risk than others, requiring thoughtful decisions about what to share and with whom. 

Applications that use cost engineering methodologies without exposing sensitive underlying data provide an easier starting point. These allow other departments to benefit from cost engineering expertise without necessarily seeing all the detailed calculations behind it.