In the first installment of this series, we explored the basics of “master data” and what it is. This time we turn our attention to the principles and technologies of Master Data Management (MDM) and why they are so important for B2B2C enterprise commerce and transformation initiatives.
Remember that, in the last post, we stressed the concept that master data is “used across multiple business processes.” Think about a product definition or a customer record. These contain master data about a customer or a product that will be accessed, analyzed, and, possibly created and/or manipulated by multiple enterprise systems. A product record may be found in systems as diverse as an ERP, PLM or an E-Commerce system as part of a catalog. Customer records may be found in a CRM, a billing system, or any number of other systems that rely on customer data. Without a method to standardize, de-duplicate, and centralize this data, each business process may be using different, and potentially out-of-date or incorrect, master data.
And, the problem doesn’t end with business processes. It crosses organizational lines. One department or geographic business unit within a large multi-national organization may implement MDM technologies and processes wholly within its boundaries, which is a great start, but without an enterprise-wide strategy for MDM, a business may find itself with multiple silos of master data, each, potentially, with conflicting information.
MDM exists in order to ensure that all critical business processes rely on the same high-quality information about products, customers, suppliers and more. (Refer back to the previous post for other important domains of master data.) Too often an enterprise will jump straight to technology to solve an information management problem, and the proper technology to meet specific use cases is necessary, but, we shouldn’t ignore the importance of the “P’s” of MDM: “People” and “Processes.” We will take up the “technologies” in the next post.
People – The success of any MDM project begins with the team that an enterprise assembles to implement and support the initiative. That team should consist of:
- An Executive champion (optimally one who is completely “bought in” on the initiative)
- Stakeholders representing every business unit with an interest in the data being mastered, including, but, importantly, not limited to IT
- A data governance team with clear and specific roles and tasks assigned
- Experienced MDM consultants to assist the team with strategy, implementation, and change management
It’s important that this team is aligned towards the company’s information management needs and balanced between the demands of the business and IT practices.
Processes – To achieve a single view of any domain of master data, there are 3 necessary business processes that must be addressed:
- Data Consolidation (Acquire) – Master data exists in multiple sources as we discussed earlier (ERP, PLM, CRM, etc). In order to centralize that master data, the necessary records and attributes must be located and migrated into an operational hub. This process can be carried out via multiple different techniques (SOAP, REST, JMS, GDSN, Manual Import, etc.) based on the source system, but, might typically during the process of Extract, Transform, and Load (ETL), the data will be validated, normalized, and classified according to predetermined standards and data models to ensure the highest quality of data.
- Data Federation and Enhancement (Manage) – Creating a single view of master data records requires a tool or algorithm that will match and link records from multiple sources into a single record that can be accessed by all relevant business processes. There are different techniques to achieve this based on the data domain. For instance, the process might rely on most recent attribute changes or maybe on a trusted system. Once multiple master data records are reconciled into a single entity, business processes can be initiated to verify the validity of the data and information, potentially from reference data or consolidation from other systems, can be added to the master record to create a more complete entity. A customer or supplier entity might include details from billing systems, a CRM or even include hierarchical or other related information. A product record may be linked to digital assets like videos, images, or product reviews. For many data domains, it is important for this process to be collaborative. In the case of a product record, multiple teams might need to create new information that should necessarily be added to the master record – and might require certain governance, or approval, workflows to ensure that the record is being properly enhanced.
- Data Propagation (Publish and Syndicate) – Once a data entity has been created through consolidation and enhanced, it is necessary to send that most up-to-date information to any system that might require it. In today’s omnichannel world that could mean sending the information back to source systems (and in some cases, it might override information in that system) or it could mean publishing to commerce channels like print, web or mobile – or even might require syndication with trading partners via GDSN or other methods. One of the key benefits of mastering data in a central location is the ability to assure that all necessary consuming channels will receive the same info, thereby ensuring that internal teams, customers, and consumers, alike, will view the same accurate data on any device and on any channel.
Of course, each of these processes requires well-planned data governance and change management strategies, two topics we will discuss in more detail in a later installment of this series.