Maturity in MDM: data silos versus multi domain approach
Master data management (MDM) supports business and IT to work together to ensure the uniformity, accuracy, stewardship, semantic consistency, and accountability of shared master data assets. Customer data and product data are the best-known examples, but many types of data fit into the same category. Strategic business decisions are made on data, the quality of this data, especially the master data needs to be 100%. For example, a business report on the sales figures of a specific product to customers in a particular region that contains multiple countries and currencies relies on various master data domains – product, customer, location, and reference data.
Master data is the uniform set of data elements that describe an organisation’s core entities, such as customers, products, suppliers, locations, hierarchies, and the chart of accounts. Important decisions within the organization depend on the quality of its master data and cadence around its management.
Piece of Cake
MDM solutions initially focused on either customer or product data, which are common to virtually any organization. Customer data tend to be high in volume, especially in telco, retail banking, and retail, but are relatively simple. Product data then tend to be less voluminous, but more elaborate and often with complex classification hierarchies. While we can’t always have our cake and eat it as well in most scenarios, there are many advantages if the management of master data is unambiguously recorded and managed like uniform cuts of cake. This concerns roles, responsibilities, functions, processes, workflows, and data rules. Putting all the pieces together so that you get the full benefits of a unified view.
When we look at MDM’s maturity in the market, it all started with implementing single-domain MDM solutions like a Customer Data Integration (CDI) platform or a Product Information Management (PIM) platform. This led to double costs in implementation, licensing and support. But even more so, these siloed solutions needed integration with another solution to share data and its relationships. Not the best way to improve data quality, data governance, and not the most ideal for business users and data stewards being confronted with working with different applications and UIs. This application architecture does not correctly support what MDM is about, sharing data assets and managing the relationships between them. This is also explained by Bob Eve, Senior Data Management Strategist, TIBCO Software, in his article “The need to become a next-generation data organisation – and leaving the silos behind”
Managing master data needs an integrated approach for data quality and data governance. And preferably within one central solution (think the full cake), so that data across multiple domains no longer needs to be exchanged through an integration solution. A change in the customer data model with an effect on product can be directly validated with the product data model.
Ingredients for Success
In an earlier article, we discussed best practices in implementing MDM as part of our customers’ organizations’ digital transformation strategy. While these best-practices can sound intimidating, a recipe for success is always about thinking big and starting small. You’ll first identify what business problem you want to solve (“I am hungry!”), following which the multi-domain idea falls in place perfectly (“I want a slice of cake”). You start looking for all the necessary pieces to put together your MDM strategy (“cream, cake, blueberries.”)
In practice, we see our customers start with the first release of a domain (such as customer). As soon as the first business benefits with a Minimum Viable Product have been achieved, expand this domain and start parallel with a new domain (such as product). This approach paves the way for a scalable and accelerable journey.
An excellent example of this is the rollout at a Dutch supermarket chain, Jumbo, and its food shop/restaurant chain, La Place. Before the launch of a loyalty card, it was a precondition to put the customer’s data in order. After the first implementation of customer MDM, a start was made with storing ingredient data of the products in the food shops/restaurants of La Place. Jumbo’s product data will soon also be included in the central MDM system. The various data models are linked so that cross-domain data insight is possible, and the governance of the multiple components can be streamlined.
La Place and Jumbo’s elegant solution is leveraging TIBCO EBX; a multi-domain platform solution acknowledged as a leader in Gartner’s MDM quadrant. Leave a comment if you want to find out more or watch this video to learn more about our unified data management approach.
