You Are Still in the First Minute of the First Day of the MDM Revolution, Are You Ticking?

What is Master Data?

MDMAny data that has to be maintained as a single version across the organization can be classified as a master data. Master data is not just limited to customer and product data. Transactional data, such as sales data, can be classified as master data if the business need is to maintain a single version of sales transaction.

Master data is generally aligned to its respective domain, such as product master, customer master, supplier master, account master, address master, asset master etc.  Multiple master data domains can also co-exist in a single environment by means of a Multi-Domain Master Data Management system.

Why Master Data?

In any organization which has multiple lines of business (LOB), the general tendency is to operate independently within the boundary of each line of business. Each LOB tends to install and maintain its own solutions with little or no degree of data sharing. Each of such independent solution tries to manage its relevant entities in an isolated way. Such an isolated model often causes several organizational problems, not only additional costs but also customer dissatisfaction.

For instance, a specific customer by name William can exist in multiple systems with variations. If all of these individual entities are not ‘mastered’ into a single best version of the truth, the results might vary from customer annoyance to loss of potential sales leads. With the explosive amount of data surging from social media, the chances of getting duplicate or incorrect entities are much higher than ever before. This makes the master data an absolute necessity in any organization. 

What is Master Data Management (MDM)?

Master Data Management is the technical/business process through which the data is mastered across several sources and destinations. This generally includes:

  • Maintains a single version of the master data (a.k.a. best version of the truth) based on the available input from different sources by incorporating attribute level trust rules
  • Provides a configurable trust model to validate and assess the reliability of the data based on the input source
  • Ensures data quality by cleansing the input data before processing
  • Removes duplicate data that originates across the same or different input sources
  • Matches and merges relevant data based on administrable rules to form the best and single version of the data
  • Maintains hierarchal information that links multiple master entities, such as the hierarchical link between customers and customer addresses
  • Provides master data versions that are based on timelines. That is, how did the specific master data entity look like, couple of years ago? Example: If the sales team wants to know where a particular customer resided last year, master data versions that are based on timelines will help
  • Provides real time data access capability though an “easy to use and easy to implement” user interface for data stewards and an API interface for external applications
  • Tighter workflow integration (Business Process Management, BPM) to publish the master data
  • Provides role-based access with attribute level granularity for security that can coalesce data encryption and masking capabilities
  • Offers customizable options to fine tune the data before and after each stage of data processing, until the best version of the truth is consolidated
  • Incorporates smart search capabilities for data stewards to search across different master data entities
  • Provides a 360 degree view of master data entity to help the business users

ACID Test

A perfect MDM system must be atomic, consistent, isolated, and durable. And, of course, Informatica MDM precisely provides all these capabilities.

Atomic – Each data attribute should be mastered separately, provided with options to define attribute level rules specific to each source system

Consistent – At any point of time, all data must be maintained in a consistent state to provide the best version of the data, irrespective of any background and foreground loads and processes

Isolated – Though a best version of the data is derived from multiple sources, data should be maintained in isolation so that a particular version of the master data from a specific source can be physically unwound to form a new version of the master data

Durable – Changes to the master data, irrespective of whether the change makes it to the best version of the data or not, must be permanent in the system for tracking and data stewardship

Informatica MDM is obviously the best in class, and possesses all the features to surpass the ACID test. Hence, it is no surprise for Informatica MDM to be positioned as a market leader in the MDM domain (Gartner Magic Quadrant 2014) for the fifth consecutive time.

Comments

  • No we are not… if that was the case there would have been banking frauds. Banks are the MOST secure in their data custody. So I find this a bit ‘sensational’. MDM is all about ‘single source of truth’ and practices in force are ensuring that. What happens in this field will not really be MDM but more on VMDM lines