There are basically two approaches to introducing MDM software. We recommend not starting with all master data domains - customers, suppliers, materials, business partners, etc. - but with just one domain to begin with. There are various options for the selection. You can tackle the domain with the biggest quality problems first or the one that is either the most important for the company or promises the quickest success. This is an individual decision. After starting with one domain, the other domains are expanded.
Success factors: data governance and process optimization
Our experience has shown that management support, structured and targeted data governance and process optimization are among the success factors for the introduction of MDM software. This is the only way to ensure that the importance of high master data quality for business and cost development is understood by all employees. Internal company guidelines for handling data are absolutely essential. Data governance defines standardized rules, processes and responsibilities for data entry, release and maintenance as well as data quality KPIs. Not only the core processes in the company, such as purchasing, production or sales, need to be taken into account. Master data processes relating to the creation, maintenance or deletion of data must also be optimized. The implementation of an MDM solution interferes with traditional structures, processes and "territories". Accompanying change management is therefore one of the success factors for turning those affected into participants and "taking them along" into the new world. Finally, a professional software solution can only ever provide support. IT support can only be provided once processes and authorizations for data maintenance and release have been clearly defined.
Use a multi-domain solution
Regardless of our recommendation to start with one domain, companies should use a multi-domain MDM solution. This is a master data solution that covers several master data domains and centralizes the entire master data management in one platform. This opens up new perspectives on the business process. Company-wide correlations and interactions become visible - and thus often a considerable savings potential in terms of time and costs. A multi-domain MDM thus creates the "one truth" for different master data domains across the entire business process.
In addition to integrated company-wide data management, multi-domain MDM systems with data quality rules and lifecycle processes can also support data governance, i.e. uniform and binding framework conditions, workflows and responsibilities for handling, maintaining and distributing data. When using various single-domain data silos, it is naturally difficult to ensure company-wide compliance with defined standards. If, on the other hand, there is only one source for master data, users have considerably less autonomy in developing definitions and rules for data, as the cross-domain data architecture is binding and transparent. The result: effective governance principles and cross-functional collaboration between departments. Both together lead to greater process efficiency and better resource allocation.
Asking the right questions
In addition to observing the success factors described above, it is important to ask the right questions in the first place. Experience from numerous customer projects has shown us this. These include the following questions: What does master data mean in our context? How is this master data defined? Which master data needs to be harmonized and which should be transferred initially? What is global or local master data? What overlaps are there between this master data and the existing data pools? What should the target processes look like? Which systems should be connected?
As a result, answering these questions by implementing a master data management solution leads to support for data governance. As all systems use the same version of the master data, the data quality "automatically" improves and the "correct" data is always available on a daily basis. Lean processes without redundant, manual data entry in the various systems - and the associated coordination effort between departments - reduce complexity and cut costs.
Define investment and change processes
Based on the processes and data governance guidelines defined in advance, the necessary creation and change processes are designed as part of the master data management solution implementation project. Workflow-based processes are used on the one hand to ensure data governance and on the other hand to ensure smooth cross-departmental collaboration when creating or changing a master record across several departments or systems. This means that they can also link processes across different systems. At an international company that specializes in the manufacture of glass and glass ceramics and uses zetVisions SPoT for the creation process for configurable products, the customer data and the individual product request are first recorded in the CRM system. From there, the basic data is transferred to zetVisions SPoT, where it is pre-entered as configurable SAP material. After completing all the data required by the ERP system via departments such as production and product management, the master data record is transferred to the SAP system. However, data transfer via preconfigured interfaces is not limited to the SAP world. In another application example, dealer data recorded in the SPoT system of a motorcycle and sports car manufacturer (e.g. names and addresses) is transferred to a content management system to enable customers to search for dealers on the company website. Processes also provide the necessary transparency regarding the status of the request until a master record is released, as the request is broken down into individual process steps and workflow mechanisms make it possible to see at any time which phase the process is currently in.
The implementation of MDM software also includes the integration of external services that make it possible to check the master data for completeness and consistency. Examples of this are: checking the VAT identification number, address checks and the embargo list check. These checks can be integrated via a web service.
Putting the MDM project into practice
In general, an implementation project takes place in several stages, with each phase being completed with a milestone before the next project phase begins. During a project preparation phase, the project scope is specified and a detailed project plan is drawn up. A kick-off meeting including team training brings everyone involved on board. In the subsequent target concept phase, which lays the foundation for the success of a master data project and is therefore the most important phase, individual objectives are defined in workshops and discussions; in addition, a needs-based concept is developed for the IT-supported mapping of company-specific data governance aspects, authorizations and the associated processes. The data models and interfaces should also be defined in this phase. Deciding which master data should be managed in the central system is also part of this phase. At the same time, the MDM system is installed on the system landscape. In the subsequent implementation phase, the individual specialist concepts created by zetVisions are put into practice. If necessary, an initial data transfer to the MDM tool can also be prepared here. There is also the opportunity to extensively test the settings or adaptations of the new system. During production preparation, zetVisions provides support with "go live" planning. This phase also includes staff training. They should not only be able to handle the system, but also be able to make adjustments to the system themselves (customizing). Once the project has been successfully completed, it is handed over to the zetVisions support team.
Monitor and optimize MDM implementation
Once the MDM solution has been successfully implemented, it needs to be monitored as part of a continuous improvement process. zetVisions offers suitable measures and solutions for this. In zetVisions SPoT, for example, ongoing monitoring of the implemented processes is possible via the integrated process monitoring. This can be used to identify optimization potential in processes.
The 5 phases for the successful implementation of MDM software
- Project preparation phase: Specification of the project scope, creation of the project plan, kick-off meeting including training.
- Target concept phase: Definition of individual objectives in workshops and discussions. Development of a needs-based concept for the IT-supported mapping of company-specific data governance aspects and the associated processes, installation of the master data management solution on the customer's system landscape.
- Implementation phase: Implementation of the technical concepts, customizing of the data models and processes, preparation of the initial data transfer if necessary, extensive testing and acceptance.
- Production preparation: support with go-live planning, employee training.
- Go-live phase: final project acceptance, system go-live (roll-out), handover of the project to support
Find out more about the benefits of standardized data and centralized data management in our blog post "Return on investment of MDM projects".
