7 guiding principles to build your data factory
Master data is the fuel that drives your organization forward. When you use gasoline in a diesel vehicle or use fuel of inferior quality, it harms the performance of your car. The same applies to master data. Incorrect/incomplete master data can slow down your entire organization. Therefore, investing in data quality should be considered the […]
Master data is the fuel that drives your organization forward. When you use gasoline in a diesel vehicle or use fuel of inferior quality, it harms the performance of your car. The same applies to master data. Incorrect/incomplete master data can slow down your entire organization. Therefore, investing in data quality should be considered the cornerstone of your digitalization journey.
Many organizations have a sophisticated system landscape with hundreds of separate systems like WMS, MES, ERP & planning systems that use their own master data records. Also, organizations often struggle, through organic growth or through acquisitions, to ensure a single view of truth across their data. This is due to inefficient controls, lack of governance & ownership.
Conclusion? Investing in your master data quality is as important as investing in your factories. A deep understanding of your master data records is highly required –aligning MD definitions & creating a golden record. So, are you ready to embark on the journey toward a data-driven organization? Let the 7 guiding principles below guide you to set up your data factory!
1. Find that golden record across your data
MD arose out of the need to improve the consistency and quality of key data assets, such as product, customer, and supplier. At the start – when companies were creating master data records for the first time – enterprises usually had one or few systems of records. The system landscapes have evolved rapidly, which has grown the need to create a single source of truth (SSOT). It is no longer manageable to maintain your master data in each separate system. Aligning master data definitions across departments and systems and finding that golden record is the first key step when setting up your MD management processes.
2. Take an E2E data approach
In modern supply chains, it is important to have end-to-end visibility over your product flows, starting with an E2E data approach. Before a product launch, it is required that the production facility creates a new item with a recipe and BOM, and the sales department creates a sales item. However, for successful data-driven organizations, the set-up of new data should be a combined effort over different departments making sure that the complete E2E flow from supplier to customer is covered.
3. Invest in analytical skills
We see a shift in the profiles that need to be involved in master data management. Traditionally master data was maintained by a business expert who has strong know-how of business processes. However, as we grow towards data-driven organizations, also other profiles are needed to ensure your data is of the best quality. First, you need someone with administrative skills who acts as a data steward guiding users through the workflow for change approvals & performing data maintenance tasks. Second, you need to embed advanced analytical skills within your data team. Data configurators will analyze transactional data within your organization and check if the master data entries are still matching reality. Both profiles often need to liaise with business experts to ensure that data entries are matching business expectations.
4. Build a data organization
The setup of your data organization is one of the hardest exercises. It’s about finding the balance between the centralization and decentralization of your master data team. On the one hand, you would like to decentralize so your data team keeps strong linkages with each business unit. On the other hand, you would like to centralize your data team, so you can maintain your golden records centrally across different departments & build an E2E data approach. In modern data-driven organizations, you would like to go for a hybrid solution, sustaining strong business linkages & also centralizing a part of the data team to maintain your master data centrally.
5. Challenge the periodic reviews
During the last few decades, master data records were only reviewed yearly in the best case, worst case never. During these reviews, the data entries are checked against business rules & business reality. However, as our world is rapidly changing, we need to challenge periodic reviews a bit more. Our data needs to be constantly in line with business reality. Lead times in your distribution network are changing significantly over time, and during your daily operations, you need to rely on quality data to make sure you can deliver on time to your customer. Therefore, we need to continuously analyze our transactional data and match them with the entries in the master data records.
6. Embed quality
Monitoring the quality of master data is extremely important because it steers your daily decision-making and operations. It is as important as monitoring the control panel of your assets in your factory. Missing a quality error can cause huge breakdowns in business process output.
You need to avoid the cost of making incorrect decisions by investing in your data quality. Big data analytics & AI tools can help you monitor/analyze your data quality across different systems, which goes beyond the check whether a field has an entry. Data quality checks become sophisticated rules checking the content of your data fields versus the applied business rules. However, data quality is not only about tooling, it needs to be embedded in your entire organization. You need to introduce the mindset that data quality is the responsibility of each individual in the organization.
7. Leverage best-of-breed solutions
Within the scope of master data management tools, we see the trend toward a move to best-of-breed solutions instead of all-in-one packages. Software vendors are often specialized in one certain topic of MDM, such as workflow management, master data governance & analytical checks. By choosing the most suitable vendor for each of these domains, you can better leverage your MDM strategy & tweak your tooling according to your specific needs.
(*)As featured in our #bluecruxtalks Webinar session of May on Master Data Factories: building master data factories, together
So, let’s agree upon the fact that investing in master data is as important as investing in your factories, and streamlined MDM processes are the cornerstone of your digitalization journey. It has become extremely important in modern data-driven organizations and should not be ignored, as centrally owned master data makes it easier to step into a system transformation journey, and quality master data avoids the costly breakdown of your process outputs.
That’s why these seven key principles are a good basis to start rethinking your own company’s MDM processes. Only with this mindset the value & importance of master data in a digital organization is correctly perceived. In case you would like to know more about Bluecrux master data factory framework, or you want to discuss your own MDM processes, get in touch and let’s win the digital race together!