Chapter 8. Master Data Management (MDM)

Master data management (MDM) is a critical part of every organization. To ensure optimum business processes, every organization must establish policies and procedures to create and maintain a singular dataset in the organization. This dataset usually contains customer details (for example, name, phone number, address, and any other supplemental information about the customer). Organizations can maintain data integrity by implementing the appropriate policies. Healthy data usually wins the trust of internal users and increases adoption of any CRM system, which can help companies make more informed decisions. Unhealthy data usually results from not having appropriate business processes to maintain the data, and so data silos are created and these prevent the integration of systems. This chapter discusses the importance of MDM and how it and emerging best practices are vital to your organization.

What Is MDM?

Master data management (MDM) refers to the policies and procedures that are implemented within the CRM system. The objective of MDM is to provide a process for data collection, consolidation, and integration between multiple systems. MDM also enables an organization to monitor and generate automated reports based on required business processes. This is usually a recursive process that evolves over time with growing company needs.

To understand MDM, you must first understand the various data structures:

• Unstructured: Any form of available data (in emails, articles, or any other documents).

• Transactional: An aggregation of any data captured during the course of the transaction.

• Reference data: All the related data that describes the core recordset. This usually contains any supplemental information that can be used for analysis (for example, list of countries, products, industries).

• Metadata: A formal data structure in a formal database combined with documents and definitions on the data.

• Hierarchical: A data structure that has layers and steps, a rudimentary structure that is defined in a tree structure (essential to create an MDM).

• Master: Subject areas, domain areas, and entity type.

Why MDM Is Important

MDM enables companies to increase internal efficiencies and drastically reduce operational cost and complexity. Thus, MDM reduces manual data entry and analyses and so improves the quality of the data. Healthy data allows organizations to make better decisions, giving the company a competitive edge over others.

The MDM process is broken down into the following key components (see Figure 8.1):

• Understanding the current data: This phase consists of profiling a segment of data, evaluating the attributes, and researching the datasets. For example, if the company directive is to ensure that the CRM data has the appropriate industry selected for the customer, the master data team evaluates the pick-list values for the industry and identifies any discrepancies in the data.

• Data cleansing: A series of sequential procedures will be followed during this phase. Data cleansing usually comprises four different parts:

• Correcting the format of the data (usually for phone numbers, ZIP codes)

• Content evaluation, which surfaces the inconsistent datasets

• Autocorrection phase, which can retrieve data from various external sources (SIC code, Hoover’s reports, LinkedIn, and more) and update the CRM system

• Manual correction

• Data enhancement: Adding any supplemental information that may prove useful in the future.

• Monitoring: Random sampling of data and ensuring the health of such.

• Reporting: Building key performance indicators (KPIs) and monitoring the general health of the data.

Figure 8.1 Master data life cycle.

MDM is often used to ensure that an organization does not duplicate master data in different operations, a common occurrence in a complex and large organization. In CRM, it is important to maintain one recordset for one customer to avoid multiple copies and to have a consolidated view of the information across various systems. For example, you don’t want an existing customer to be contacted by a salesperson soliciting the same service or product. (Such a mistake might occur if, for example, the sales department and the customer relationship department information are not synced).

This problem may be amplified by mergers and acquisitions. Consider the example of World Savings being taken over by Wachovia being taken over by Wells Fargo. This entire takeover process would have created three (World Savings, Wachovia, and Wells Fargo) sets of customer data. When the three companies merge, the integrators need to ensure that the customer records are merged as the organization is merging. Another potential problem that may happen during this merger is the redundancies of products and services offered by the organization; typically, this is put on a backburner and addressed at a later stage. A consolidated view of customer contact information, along with the products and services used, can enable the business analyst to provide better options for the management to choose from. MDM can help reduce serious operational problems and increase operational efficiency, which will then lead to smarter business decisions.

MDM tools provide a suitable file system structure and an adaptive data warehouse system, an operational database enabled with data mining and analysis capabilities, and they allow for automated reporting and exception-based reporting. Document all of this, and diagram it for easier understanding. You can use any available modeling methodology acceptable to the organization.

MDM Tools

SQL Server 2008 R2 Master Data Services (see Figure 8.2) enables organizations to define business processes and the data processes for the SQL server. Thus, organizations can implement human-to-human, human-to-system, system-to-human, and system-to-system workflows.

Figure 8.2 Master Data Services home page for SQL Server.

Summary

MDM enables an organization to integrate various information systems into one consolidated and coherent system. A healthy CRM system results in high user adoption. Salespeople are attracted to a single system with a healthy dataset. If the dataset is poor, you might be caught in a downward spiral that may lead to internal users losing trust in the quality of the data. In most organizations, MDM is a recursive process. The salespeople are usually the best source of the information and are the biggest consumers of the information.

It is important to implement a process to extract the information and to then store it in a centralized location. A good MDM strategy allows for communication with different functional business roles in an organization and keeps them informed of the effort of the other roles in the organization. It also implements best practices that enforce corporate standards. Implementing MDM in an organization will empower the company to develop and deploy an automated information management solution.