Master Data Management — Part 3: How to Achieve the Ultimate Goal of MDM

This is Part 3 of our mini series 'Master Data Management — What It Is and Why You Need It'

Terminology:

• Single Source of Truth (SSOT): A data storage principle to always source a particular piece of information from one place.
• Golden Record: A data record that captures all the necessary information we need without redundancies.

The ultimate goal of MDM is to have an SSOT so that your data is in one place, and a golden record so that each time you access it, you know that's the latest and most accurate version of your organization's data.

There are many ways to achieve an SSOT and a golden record, but we've found the following to be the best approach.

Step One — Scope and Proof of Concept

As a first step, you need to understand all the data requirements related to the scope of your project. You might want to start with a small Proof of Concept (POC) to demonstrate its feasibility.

Step Two — Data Analysis

Next, you will need to undertake extensive data analysis in order to map the relevant master data from all sources. This is important to understand the quality of your data and determine how to unify it for your goals.

Step Three — Technical Design

The next step is to figure out the design of your MDM solution. This includes figuring out the technical design and the Extract-Load-Transform (ELT) process, which is all about extracting the data from multiple sources and loading it into a single source, usually a data warehouse.

Before getting into MDM implementation, you have to establish how your business hopes to use your data. Do you just need clean, coordinated data for reporting (analytical MDM), or do you want validated data to help inform business processes in real time (operational MDM)? While most businesses want to work toward a hybrid MDM approach, analytical MDM requires a significantly less complex implementation of its architecture than operational MDM. 

Step Four — Implementation

Now it's time for implementation, which is where your SSOT is created and made available to your organization. You now have an effective Master Data Management system. Now of course that is all easier said than done, and there is a list of key technical factors to consider. 

Key Technical Considerations
The following technical considerations are simplified. If you would like to receive a complimentary copy of specific implementation techniques that we have developed based on our research, grab it here:

Download Implementation Techniques

Analytical MDM

• Avoid data transformations and data grouping directly in the dashboard, manage data operations in the data warehouse.
• Create a data warehouse where the dirty, the clean and the golden records are organized.
• Allow only data from official data sets in the data warehouse to be used for analytics and dashboards.
• Define strict roles and access policies for different layers in the data warehouse
• Define metrics, mappings, groups and technical parameters across all departments of your organization.
• Use tags to group related master data information.

Operational MDM

• Prevent data duplication in an operational system by not using free text fields – use data validation instruments, if available.
• Exclude certain types of data, create valid ranges.
• The use of CDC (change data capture) to detect overwritten and deleted data enables you to show history of MDM attributes.
• Use an algorithm (fuzzy matching, clustering, etc.) to merge duplicate data entries and feed information back to systems.

We hope this has given you a better understanding of MDM, and why you should care about it. Master data management is critical to the success of any data-driven organization and its ability to make accurate decisions based on data.

If you have any questions or would like to know more about MDM, please do not hesitate to contact us, or receive a complimentary copy of our specific MDM implementation techniques  — we would be more than happy to help.​

Read more:
Part 1: What Is MDM?
Part 2: Why Is MDM Important? ​