In a world obsessed with data holding and storage, Master Data Management is the need of the hour. It’s no longer a ‘thing to think about in the future’, but an urgent, necessary process that businesses of all sizes need to make sense of their data.

In this value-driven comprehensive master data management guide, we will help you define & implement a master data management strategy and address key challenges.

The guide is categorized into two parts:

    1. Basics of Master Data Management: Understand the what, why, and how of MDM processes. This part acts as a refresher for professionals and a quick guide for beginners.
    2. Implementing a Master Data Management Initiative: This segment covers everything on MDM strategies, hidden MDM costs to look out for, challenges, and issues to prepare for.


Let’s dig in!

Basics Of Master Data Management

Data by itself is meaningless.

In order to give data meaning, you have to clean, organize, and structure it in a way that allows for easy access and insights.

For instance, if you’ve got two decades’ worth of data but 40% of it is redundant, outdated, or dirty, you cannot retrieve reliable or accurate insights from the data! 

Failing to master your organization’s data would result in costly expenses, poor insights, and failed business strategies.

what is MDM Winpure

What is master data management? 

Put simply, Master data management or MDM is a type of data management architecture, governed by data quality standards and practices, designed to help businesses make sense of their data.

Think of master data management as the plumbing in your house. There’s a reservoir (a data hub) to which many pipes and lines are connected (software, tools, or dashboards). The plumbing is designed in a way that allows water to flow smoothly into the taps of your kitchen, bathroom, and other areas where water is required.

Similarly, MDM is allows different departments in your organization to access data to perform key tasks like:

  • Planning marketing campaigns
  • Reviewing sales performance 
  • Budgeting and forecasting
  • Hiring and cost-cutting
  • New business or product line launches
  • New customer experience initiatives

Without access to accurate records, every department in your organization will struggle with poor insights, inefficient work processes, and very angry customers who may even sue you for sending unsolicited marketing materials. MDM is now a critical asset for a data-driven and data-informed business.


A Quick History of MDM

Master Data Management originated in the 1990s to deal with lots of disjointed master data. For example, schools would find the same student represented in different systems for admission, registration, textbook sales, grading, career counseling, and alumni services. The student would engage with various departments for different needs, each place duplicated the same types of attributes, causing inefficiency and likelihood of error. Schools and other businesses looked to MDM to clean this overlapping raw data for use in any department.

Before computerized systems, MDM meant setting up, using, and maintaining a folder system in a filing cabinet combined with using interdepartmental mail. For instance, hospital volunteers called candy stripers would physically deliver patient data from one department (e.g., radiology) to another (e.g., a physician’s office filing cabinet). Some of the same data from a patient’s folder in a doctor’s office would go to a nursing station or admissions (usually through some standardized forms). This required workers to sift through paper and wait for information to be delivered.

master data management history

Today’s top MDM systems use software and networking advances, people, and data governance processes to get work done more efficiently. Rather than paper, current MDM controls and operates on Master Data and Reference Data explained in the next sections.

What Is Master Data In Simple Words?

Master data is the core data that is needed to run operations in a business. Master data is a set of consistent and uniform identifiers. The identifiers can be prospects, customers, assets, suppliers, products, account charts, sites, hierarchies, locations, etc.

Typically, master data breaks down into four domains:

  • Customers
  • Products
  • Locations
  • Other

Master data tends to be harder to visualize as it describes core data assets distributed across many different databases. This associated information belongs to specific entities necessary for a business to operate and grow.

Each of the four domains can be further divided into subsections:

  • Customers –>  Company name, address, email, and phone.
  • Products –>  The good or service in use and the ingredient or part to make it.
  • Location –>  The region where the head office is located.
  • Other –>  Calendar of company events, licenses.

Likewise, these subsections can be split into further subcategories.

Any level of master data comprises a single source of truth or golden copy that remains consistent and non-duplicated. The power of master data lies in that it can be copied but not altered in more than one place, a hub.

Hubs containing stored and accessible master data come in different models:

  • Registry –>  checks the MDM hub to correct any contradictions among multiple sources.
  • Consolidated –> gathers master data from multiple systems into the hub where it is sourced.
  • Coexisted –> records duplicate master data with the master data in the hub and the duplicate in each source system.
  • Transaction-based –> enhances the data through algorithms ensuring the data is accessible in the source system.

Keep in mind that master data, in any hub, is just information. To turn this data into insights and analytics, you would need to implement MDM governed procedures and processes to make sense of the data.

What Is Reference Data?

Consider reference data, a type of master data that functions as a “Rosetta Stone” for enterprise and departmental applications to understand and work with one another. Typically, other data in a database or external information needed to do business constitutes reference data.

Another example of reference data is with the USPS. They list the state, as pull-down menu values on a form. See below:USPS zip code

What’s The Difference between Data management, MDM & EDM?

Data management covers a broader concept than enterprise data management or MDM. It describes “a comprehensive collection and hierarchy of practices, concepts, procedures, processes, and a wide range of accompanying systems that allow for an organization to gain control of its data resources. This spectrum includes business strategies, databases, business intelligence, and MDM, among others. Data management’s importance lies in its tremendous monetary and social value, as the Harvard Business Review puts it.

Likewise, MDM does not work alone within an enterprise data management strategy. MDM requires good data quality and reference data management towards providing master data. The enterprise data architecture and privacy and security require appropriate access and control for master data, and all the other enterprise data. MDM cannot work alone as an enterprise data management solution.

In the 2010s, data management became particularly necessary. More significant and frequent data breaches disrupted banks, supermarkets, governments, and a hospital. Also, ransomware gained traction, with increasing monetary demands and dangers to privacy. Governments recognized the high financial and political outcomes from data. In response, legislation from Europe (the GDPR) and California (the CCPA) order companies to protect an individual’s privacy. Companies and individuals recognize that they need data management to prevent fines and mishaps and to take advantage of new opportunities like AI and machine learning.

What Is Master Data Management Architecture?

Master Data Management architecture is a way to record and document a business’ data assets. It maps how data flows through the organisation’s systems and how the data is organised to serve the business

MDM describes one type of data management architecture designed to connect many different data sources, easier. Since MDM keeps shared data standardized according to Master Data Governance, data is better controlled and more accessible. As a result, MDM gives companies a huge advantage to use all sorts of technologies and process improvements quickly.

Is Enterprise Data Management the Same as Master Data Management?

Using the house metaphor for enterprise data management provides another way to think of MDM. A building has many different systems besides plumbing. Water gets heated by gas or electricity to provide hot water. Most likely different companies provide gas and water. But you will not get hot water alone. You need the gas for plumbing and the house to be functional.

While MDM and enterprise data management make information more useful throughout the company, they have slightly different meanings.

Enterprise data management describes a “holistic framework comprising the people, processes, and technology that optimizes data from a variety of different sources, then makes it available when and where it’s needed.” Think of Enterprise Data Management as a type of blueprint, like a house. Within enterprise data management, MDM describes a specific platform that reduces transaction costs, integrates data, standardizes critical data, and simplifies data architectures across the company so that the entire business can function.

MDM cannot work alone in Enterprise Data Management without other data management components.

mdm vs edm
Difference between data management, master data management and enterprise data management

How to implement the MDM initiative?

A master data management initiative consists of three parts:

1). Identify a Business Problem: An MDM process is either tied to a business goal or is initiated to solve a business problem. It’s necessary to have a clear vision of what you’re trying to achieve with an MDM activity.

2). A Master Data Management Strategy: An MDM strategy will lay the groundwork for the activity and will help you identify steps you need to prevent chances of failure.

Related: Here’s everything you need to know about what is a master data management strategy and why it is important.

3). Anticipating Challenges & Hidden Costs: What are the potential challenges you need to watch out for and what to do to avoid failure.

Let’s elaborate on these components in detail.

Identify Business Problems

Generally, companies don’t worry about master data management unless a crucial business process is being affected or there’s a business problem that needs to be resolved.

For instance, most companies would not bother with consolidated views unless they have to get homogenous data for a marketing initiative or targeted advertising for a new product line.

In such cases, the business requires a consolidated view of clean, duplicate-free data from multiple sources.

Data Health Checklist
A Complete Data Health Checklist

If they choose to ignore dirty data and continue with blatant marketing & advertising, it will result in problems like:

  • Loss of customers caused by poor targeting
  • Loss of brand reputation by ineffective marketing
  • Direct financial losses caused by poor planning
  • Internal conflicts and employee lay-offs
  • Irked stakeholders & plummeting shares

The hit to business efficiency & productivity is severe. Hence, it’s necessary to tie a business goal or problem resolution to an MDM initiative so the impact can be measured.

Create a Master Data Management Strategy

Organizations now realize they need a reliable, updated, and accurate view of their data. In fact, many companies are beginning to implement MDM processes to get a singular view of a customer’s journey. However,  an MDM initiative is easier said than done. The chances of failure are high.

Gartner estimates around 90% of businesses fail when implementing an MDM strategy.

So how to launch an MDM initiative with lowered failure risks?

By creating a strategy that involves C-level buy-ins to KPI measurements.

Master data management steps

Step 1: Get C-level Buy-in for MDM Initiatives  

Decision-makers and C-suite executives are often hesitant when it comes to addressing data issues. They are risk-averse and any initiative that threatens to remove, delete, or even move data is considered a significant risk.

Therefore, getting the buy-in of decision-makers is the first step to a successful MDM project plan.

Here’s how you can do that.

    • Start with a small pilot project. Show how a consolidated view of one department’s data sources helped solve a business problem. For instance, organizing customer data for targeted advertising. Demonstrate how the removing of redundant or duplicate data, led to lower advertising costs & higher ROI. After a pilot project is successful, you can then use this as a case study to demonstrate how poor data quality or poor data management is affecting business outcomes.
    • Involve business stakeholders. They can best articulate how data is affecting their operations and needs. As in the example above, involving your marketing managers is the best way to demonstrate how an MDM strategy can fix a critical marketing problem.
    • Record real business value. Master data initiatives are built around business value. Connect your MDM initiative with a business driver and use the KPIs your executives use to measure the success of this initiative. Some examples of real-business value or KPIs that you can use to convince C-level executives:
            1. Improve customer experience rate
            2. Improve marketing campaign conversion rates
            3. Improve organizational efficiency
            4. Automate regulatory reporting
            5. Deliver consistent customer experience 
            6. Streamline mergers and acquisitions 

Demonstrating these KPIs or resolution of business problems will facilitate a quicker C-level buy-in; which generally can take months.

Step 2: Plan Your MDM Roadmap

When you’re trying to implement an MDM initiative, it pays to know exactly where you are today and where you need to be a few months from now.

How do you know where you are today?

By doing a health check on your data across the organization. Questions to ask:

  • How is the data captured, stored, and managed?
  • How many sources of data are connected to your central database 
  • How well-governed is your data? Do you have data governance standards implemented?
  • How much of the data is structured, semi-structured, or unstructured?
  • How much do you spend manually fixing data as opposed to automating data management?
  • How are your teams aligned with each other when it comes to accessing & processing data?
  • Do you have regular internal conflicts between IT and business users?

And the most important question:

  • What is the state of your data quality? Is your data timely, complete, accurate, unique and follows standardization rules?

Your answers to these questions will help determine your roadmap and milestone planning. It will also be instrumental in helping you choose the right MDM tools or MDM solutions for the business.

We’ve written a whole comparison guide on the best Master Data Management Tools in 2022, have a read here.

Step 3: Define Data Owners & Key Roles

An MDM project is not just an IT activity. It is a business process, involving business users who are the owners of their data.

However, before assigning roles, it’s important to address user access rights and domain ownership.

For instance, a small business with CRM data can assign user access according to seniority; like giving master ID access only to the department head while giving view-only access to other users.

Similarly, an enterprise, like a bank, or a retail store, will need a much larger number of team members to access, amend, and manage data. In such cases, it is necessary to grant access on the basis of role and frequency of interaction with the data.

Step 4: Call for MDM Vendor RFPs

Choosing a master data management vendor is a mind-boggling task. There are hundreds of solutions in the market, each claiming a special service, feature, or technology. There are different options for similar problems and choosing the right one can make or break your MDM plans.

If you’re able to be specific on what you want to achieve up front and what features would be deal-breakers, you can narrow down the options more quickly and avoid using a solution that is not the right fit for your business.

Some key questions to ask when choosing would be:

An illustration that explains how to choose an MDM Solution
how to choose an MDM solution

Once you have answers to these questions, you can begin your search on finding the best master data management tool. Make a list of top vendors and request for RFPs. Choose your vendor based on features, specific answers to your requirements, and customer service.

An MDM strategy could take years to plan and perhaps tens of millions in quotations by vendors. WinPure hopes to help reduce that to 30-60 days for presales, and a total of 60-90 days to start using the solution.

Related Guide: Best Master Data Management Solutions In 2022

Step 5: Set Milestones and Measurable KPIs

It’s very easy to get lost in the overwhelming process of master data management, hence a robust project management plan is essential to stay on track.

Assign milestones and responsibilities. Measure KPIs and outcomes of achieving milestones. For instance, how much were you able to cut down by removing 3,000 duplicate records in a direct mail database?

Did you know you can save thousands of dollars on postal mail alone if you have clean data!

What are MDM tasks and activities?

It is a general presumption that an MDM activity is just about consolidating data to provide a golden record or a single source of truth for the business.

A master data management activity, however, does more than this. Consolidation is just part of a much bigger picture.


Some of the key tasks of an MDM activity include:

  • Identifying data sources containing information about the entity to be consolidated.
  • Creating a target or expected view of the master data model for the selected model.
  • Developing scripts or using an MDM tool to pull data from multiple sources.
  • Performing data cleansing and data profiling of the extracted data.
  • Normalizing unstructured and semi-structured data.
  • Developing merging rules for record linkage by identifying key attributes.
  • Purging duplicates or redundant data.
  • Create a master index through which entity identification and resolution can be performed.
  • Load the consolidated record into a master repository.
  • Develop automation to repeat processes periodically.

The objective of an MDM task is to create a single source of truth or a golden record which is expected to be the “clean” “master” data combined from multiple sources. This marks the end of the project and thereafter business rules are created to maintain or manage the master data.

What are the Challenges of MDM Plans?

MDM initiatives fail because most businesses do not address challenges as:

14 Key Challenges of an MDM Initiative

  1. Data quality issues were largely ignored.
  2. Selection of MDM solutions that are not agile
  3. Lack of data governance policies and business rules
  4. Focus on technology acquisition over the fulfillment of business needs.
  5. Uncoordinated MDM activities initiated in isolation
  6. Issues with scope definition and poor business process planning.
  7. Lack of 100% support from C-level executives.
  8. Incorrect prediction of costs or failing to identify hidden costs.
  9. Hiring the wrong talent to lead the project.
  10. Neglecting basic risk management protocols.
  11. Loss of data during a linkage or merging attempt.
  12. Loss of meaning when data consolidation is done in isolation by IT staff
  13. Misalignment with business scope.
  14. Failing to address third-party or external data connections

MDM is an enterprise-wide activity that must go beyond the needs of any single business function. It’s not merely a task of consolidating data sources – it is a business process that requires collaboration from multiple departments, and must address barriers to success.

What are the Hidden Costs of an MDM Project Plan?

It’s hard to apply a dollar value to the costs of an MDM project plan as it differs from company to company.

For some companies, the expenses of an MDM project can rise up to $100K/month taking into factors like software licensing, storage, implementation, talent hiring, training and maintenance, and support.

For others, it could be just $10K/month if they already have a team and are just using an affordable MDM solution to master departmental records.

MDM costs can be roughly calculated by assessing:

  • Vendor Costs – The software license costs for an average size implementation of an enterprise MDM solution range anywhere from $500,000 to $2 million
  • Hardware costs – servers, storage units, systems etc can amount up to $1million
  • Development and deployment – talent and consultancy costs can go up to $50K a month

According to an article on CIO, some large organizations will need a potential investment exceeding $10 million for MDM success.

What is the ROI of MDM?

An MDM initiative is costly, but the ROI overrides the costs. Here’s a neat breakdown of MDM ROI by Kelvin Loi, of Mastech Info Trellis.

A table showing the ROI of MDM

Details of this five-year MDM ROI projection are given here.

As you can see, an MDM process resolves business bottlenecks which in time does not only reduce expensive costs but also increases ROI.

Rethinking the MDM Concept & Going Beyond the Golden Record Vision

Traditionally, MDM is considered a technical activity that solely seeks to combine data sources, master them, and deliver a golden record master repository.

But now times have changed. Data is no longer an IT asset. It is an organizational asset that is owned by everyone in the organization.

The presumption that data has to be managed within siloed budgets and governed by only specific groups has led many MDM projects to failure.

It is critical for companies to rethink the MDM concept and choose to think beyond the Golden Record Vision.

Here’s how you can rethink MDM and ensure an MDM initiative results in success.

  • Ensuring Consistency: the structure and representation of data across the organization must be consistent. For example, if customer names, location, and the number have varied data structures (Sean vs Shawn) or (UK vs United Kingdom), then merging and eliminating those variations without assessing the impact will lead to process flaws downstream. Setting consistent standards and assessing current data models against those standards will help save you from costly mistakes and process flaws during the MDM activity.
  • Collaboration b/w IT and Business Teams: The IT team is often relied on to make strategic decisions about data governance and data management, even though they are often not the owners of data. Imagine the impact of handing over sales or customer service data to the IT team who may not be familiar with commonly used business terms or other metadata descriptions – conflicts will escalate when data is merged or purged without addressing context. Business process owners need to be part of the data consolidation process because no one else understands context, relevance, and purpose better than them
  • Data Quality Management in MDM: The first step to a successful MDM plan is implementing a data quality plan. It’s imperative that organization-wide data requirements be balanced with an organizational framework for data quality assurance and management. This includes setting data quality standards, providing data quality training to people responsible for holding, storing, or entering the data, cleansing when necessary, and consistently measuring/monitoring data quality.
  • Establishing Data Governance in MDM: Putting the right amount of data governance policies in place is critical to ensuring consistency of data models and resolving data issues on time.
  • MDM for Improved Customer Experience: MDM shouldn’t be an activity just to get consolidated views. It must be tied with a key goal of improving customer experiences. For example, if your marketing or sales team is still targeting the wrong customers with irrelevant products or campaigns, your MDM initiative has failed to optimize customer experience.

Master data management must implement best practices to enable reliable, accurate, up-to-date, and comprehensive access to information and preserve information visibility.

Related: See the benefits of master data management?

WinPure Can Help You Kickstart an MDM Pilot Project

As we said earlier, MDM is an expensive endeavor and not one that you should get into lightly. Plus, in order to get approval from C-level decision-makers, you will have to prove your case for MDM.

WinPure’s solution can help you kick off a pilot project where you can master the records of one department or one entity to determine your next steps. For enterprise organizations where an MDM project can take months to plan and get approvals, WinPure can accelerate the process faster with a test run.

On the other hand, if you’re a small or medium-level business that does not have up to $10 million to invest in MDM projects, you can use an affordable solution like WinPure to achieve your MDM goals.

Here’s everything you need to know about WinPure’s MDM solution and how it can help you achieve a single view of your data without impacting your business processes or draining your budget.

Get in touch to see how we can help you with making sense of your data through a simple, easy-to-use, affordable tool for your business.

A form only for entity resolution demo bookings



Written by Michelle Knight & Farah Kim

Michelle Knight has a background in software testing, a Master's in Library and Information Science from Simmons College, and an Association for Information Science and Technology (ASIST) award. At WinPure, she works as our Product Marketing Specialist and has a knack for explaining complicated data management topics to business people. Farah Kim is a human-centric product marketer and specializes in simplifying complex information into actionable insights for the WinPure audience. Farah holds a BS degree in Computer Science and a MA degree in Linguistics. She is fascinated with data management and aims to help businesses overcome operational inefficiencies caused by ineffective data management practices.

Share this Post

Schedule a Demo

Explore WinPure’s award-winning data quality suite packed with capabilities like:

  • Data Profiling
  • Data Cleansing & Standardization
  • Data/Fuzzy Matching
  • Data Deduplication
  • AI Entity Resolution
  • Address Verification

…. and much more!