FIVE CHALLENGES IN IMPLEMENTING MASTER DATA MANAGEMENT
At the beginning of 2021 Michiel van der Lans wrote a blog about the principle of Master Data Management (MDM). In it, he explained how an MDM application works and introduced us to how ITDS sees the master – slave relationship. In this blog, Business Consultant Noah de Groot walks us through the process of actually implementing an MDM solution. What are the pitfalls to avoid if we want to use MDM in the organisation? And what kind of challenges are there for financial-services providers?
MDM and financial services
But first, let’s take a step back. What should a bank, insurer or pension fund weigh up before even considering the implementation of MDM? In other words, what relevance will MDM have in the provision of their financial services? The answer to this questions lies in the issues that financial-services providers have to contend with.
The two most important of these are the increasing pressure of legislation and regulations that they face, and the extent to which they harvest data from their customers. This data can be valuable, provided it is of a high quality, is efficiently organised, accessible and uniformly available throughout the organisation. If these provisions are met, an organisation can manage data as an asset and thus underpin its ability to keep tailoring its products and services to the requirements of its customers. Furthermore, legislative and regulatory bodies require that banks, insurers and pension funds have all their data-management ducks in a row, so that they can comply with privacy legislation such as GDPR and Customer Due Diligence, for example. These require that data quality is properly in order, which is exactly where an MDM solution can help.
MDM as the spider in the web
Financial services providers use a variety of data warehouses, data lakes and applications to manage their relationship data. This data is often stored in fragmented form across the organisation and its various applications or databases. It can result in the same data point, such as a customer’s address for example, having two different values. An MDM application focuses on the management of Master Data. By carrying out various checks and validations on the data, according to business rules, it creates a single database with a single truth. This data is then communicated to the data sources, the so-called Systems of Record.
An MDM application can offer a data-quality solution for many financial-services organisations, but it all stands and falls with the proper implementation of that solution. When implementing MDM, an organisation typically faces five common challenges.
1. The agility of the MDM model
The type of MDM model that you adopt in your organisation can greatly impact the success of its implementation. The data model must be sufficiently agile and able to adapt to changes in complex systems. If a data model cannot do this, it will cause even more data-management problems. In selecting a data model with the required level of agility, an organisation will have to take a number of steps: create the data model; define the business rules; define the data-validation mechanisms; and define the roles and control measures.
2. Choosing the master dataset
For the effective implementation of MDM, it is essential that you choose well the elements of the data that are to be “mastered”. Different departments will often have different interests, and arriving at a standard set of data elements can lead to a lot of discussion. An insurer’s non-life department, for example, will consider a customer’s email address to be very important, while the pensions department will want to know the name of the customer’s employer. This makes it very important to establish a broadly supported agreement about which standard to use. Creating a Canonical Data Model (CDM) can help in this respect. A CDM aims to present data entities and their mutual relationships as simply as possible, thus creating a standardised model that is understandable and clear to everyone.
3. Good data governance
Given that implementing MDM is a complex process, having a mature data-governance structure in the organisation is all-the-more important. And if you are to effectively monitor the policy, activities and responsibilities, the respective roles in this structure must also be clear. Without good Data Governance, your chances of successfully implementing MDM are that much smaller.
4. Data integration
Integrating your data applications with the MDM application can take up a lot of time and capacity. Furthermore, migrating data from one application to another can introduce errors. Doing it all properly can be a time consuming process, but the correct integration of the applications will be key to underpinning the functionality of your MDM solution. And it also works both ways; an adjustment in the “master” data must also be properly communicated to the “slaves”.
5. Improving data quality
The quality of the data is a very important factor in an MDM implementation. If your data is corrupted, the master data will then also be corrupted. You’ll create one truth, but it won’t be the right one. A solution here could be to appoint data stewards in the organisation. Data stewardship guarantees the quality of the data by making sure it gets the attention it deserves. Better data quality makes for a smoother MDM implementation.
Master Data Management difficult to master? ITDS makes it easy
An MDM implementation can be a complex process. Many organisations will face challenges they won’t have foreseen, often leading to delays in the implementation process and higher costs. By taking the above-mentioned challenges in account when planning the implementation process and developing the applicable solutions, your chances of success will increase dramatically.
Want to know more about how ITDS can help? Contact Noah de Groot at firstname.lastname@example.org.