Definition of MDM
Master Data Management (MDM) can be defined as an ENTERPRISE-WIDE INFRASTRUCTURE for INTEGRATION, HARMONIZING AND MANAGING of MASTER DATA, so that information is SUPPORTIVE of BUSINESS DECISIONS to ENHANCE ORGANIZATION VALUE.
Break-up of definition
Master Data Management (MDM) can be defined as an
ENTERPRISE-WIDE A MDM initiative brings the value, if it is an enterprise-wide. The objective is to make Master Data as an enterprise asset, instead of functional level asset
INFRASTRUCTURE - This include hardware, software, people, procedures, integration services, web services, administration services etc.
For
INTEGRATING-This includes bringing Master Data from various sources into a single and common reference point, while ensuring quality and consistency
and
HARMONIZING - MDM is much more a business initiative than an IT initiative. Harmonizing includes common business rules, interpretation and processes linked to Master Data.
and
MANAGING - MDM is not only an initiative, but is an ongoing effort to create, maintain and above-all 'utilizing' the master data.
MASTER DATA - This is all the non-transactional business data. Transactional data examples are Invoice, Accounting entry, reimbursements, payment receipts, finished goods production etc. Master Data examples are Customer Master, Vendor Master, Parts Master, Product Master, Employee Master and Location Master etc..
so that information is
SUPPORTIVE- MDM has to serve a purpose and support the business objectives. MDM has a significant operational use (unlike BI, which is mostly non-operational)
of
BUSINESS DECISIONS - MDM design and implementation has to be guided by the business objectives behind the MDM initiative. You can have different MDM implementations depending upon the business objectives.
to
ENHANCE ORGANIZATION VALUE- Beyond supporting business decisions, MDM platform should provide a long term embedded value. Apart from adding to the immediate financials of an organization, it creates an enterprise level asset, which builds foundation for future growth.
Master data Management (MDM) and Customer Data Integration (CDI) are used inter-changeably
As you go through the MDM section, we will be using MDM and CDI interchangeably because of the following reasons:
- Customer means not only external customer, but also any entity like employees, business entities, sales agents etc..
- Most of the applications related to MDM are around customer master data. This is typically the start point of MDM initiatives.
- The reader can relate more to the CDI.
High Level description of How MDM works
In very simple terms, the Master Data is integrated across the source systems containing bits and pieces of the Master Data. This integration involves merging, de-duplication, standardizing, cleansing, and other transformations. This integrated data is placed in a central repository called MDM hub.
The question is on what's different from Data Warehouse? The data in MDM-CDI Hub is referred by various applications on real-time basis to fulfill their operational activities as well as for doing analysis. MDM hub actually supplies quality data to DW platform, which in-turn supports BI for all kind of data. Data Warehouse is generally used for offline applications.
The physical MDM hub is one way to implement MDM. You can also have a federated model, whereby HUB contains only the pointers to the master data, which physically is residing in the source systems. There are various architecture options in between the 'centralized physical' and 'truly federated' model.
What MDM is not
MDM is not:
- An IT initiative- It's a business initiative with IT playing a support role. However, IT is a core player as you cannot implement an MDM without an IT platform.
- An Application- MDM is not a business system application. It is a foundation platform, which provides reference data for business applications and business processes to become more effective.
- Business Intelligence- MDM is an initiative to provide single point of reference for Master Data. It serves many purposes and BI is one of its many beneficiaries.
De-Mystification Comparisons
CRM vs. MDM
Master Data Management enhances effectiveness of CRM, but is not CRM. The differences are:
- Master Data Management is essentially focused on Master Data, which could be any type of Master Data like Customer, Product and Location etc... CRM is focused on the wide-ranging processes and technology related to Customer Entity.
- CRM is subject of business management, in terms of how we relate with the customer including customer Satisfaction, Customer Value etc. Master Data Management is not the subject of Business Management, but of Data Management. Data Management helps in making business management more effective.
- CRM is essentially enabled by business systems applications (like sales force automation, Customer Contact Centre applications..), whereas MDM provide enablement to business applications, with CRM being one of them.
Business Intelligence vs. MDM
- MDM serves all stakeholders of an organization and not only BI. Business Intelligence is one of the key beneficiaries. BI also benefits MDM, as in many cases MDM rides on the infrastructure elements used by BI. The example of these infrastructure elements are- ETL tools, DBMS platforms etc...
- MDM initiative integrates the Master Data using varied integration techniques. A typical BI platform will be using ETL as standard integration technique.
- BI manages both Master and transactional data, whereby MDM is focused on Master Data.
- BI is mostly used for offline purposes, whereby MDM platform is used for online real-time application.
CDI vs. MDM
CDI (Customer Data Integration) is one among many domains of MDM. You can have Employee Information Management, Vendor Information Management, Parts Information Management etc..., in MDM. CDI is most popular of MDM domains, and typically MDM is used interchageably with CDI.
MDM vs. Data Warehouse
- Data Warehouse is a repository of all types of Data, including Master Data, Transaction Data and Historical Data etc...MDM provides a reference point to Master Data only.
- Data Warehouse mostly is used for non-operational, offline purpose (Enterprise Reporting, Data Mining, Analytics, Performance Reporting etc..). MDM is used mainly for operational and also for offline purposes. A conventional architecture is for MDM repository to feed the Master Data to Data Warehouse. Therefore MDM repository (or what you call MDM hub), becomes one of the source systems to Data Warehouse.
- Data Warehouse is mostly update on Daily basis (Refer Refresh frequency of a Data warehouse) via end of the day processing. MDM is generally updated real-time or almost real-time basis.
- Data Warehouse does not allow write-back, whereas MDM HUB is designed for write-back.
MDM vs. Business Application
MDM is a foundation platform (like a Data Warehouse), which supports the business applications and processes to be more effective. By itself, it does not perform and business function.
MDM vs. Metadata Management
- Metadata is covering all kinds of data WHEREAS MDM is concerned about non-transaction master-data only.
- Metadata is not concerned with integration and synchronization of the data (though it does support that objective) WHEREAS master data integration is core to the MDM. Technically speaking, you can have a metadata repository, without a good synchronization and integration of data, but you cannot have an MDM, without the same.
- It is not mandatory for Metadata to enforce common data standards and business rules (though it supports that objective) about data WHEREAS MDM fundamentals are based on common standards and business rules.
Domains of MDM
CDI (Customer Data Integration)
CDI is perhaps most popular domain for MDM. Customer data has unique set of challenges. CDI can be called most rewarding and also among difficult MDM initiatives. The difference between CDI and other domains of MDM is that customer data is mostly generated externally.
PIM (Product Information Management)
Even a mid-sized organization will be dealing in thousands of products, if we take into account all variations in terms of size, model, color, accessories, multi-product packages etc...PIM is unique because, its model itself keeps on changing. As product mature, it will keep on adding new attributes. In this age of 'personalization', companies try to bring various shades of their products to suit the needs of a customer. For example A Car today can have dozens of options like type of tyres (Radial, Extra Radial, Tubeless Radials..), Wheels (Simple Foundry Wheel, Alloy Steel Wheels, Special Black Painted Alloy Wheels..), Color of the car, Color of Bumpers....
One needs to maintain Master Data on the base product, add-on features and each of the thousands of permutations and combinations. By this way you can manage the inventories, prices, discount policies, product profitability etc... PIM enables Product Development and Management, as it gives a complete view of the product portfolio.
EIM (Employee Information Management)
Employee information management is perhaps one of easier domains to manage. One of the benefits of having an Employee Master Data Management, is that it enables centralization of the payroll, employee communication, User administration and security management (both IT level and Physical level) etc....
We have not seen many MDM applications around employees. This is because most of the progressive organizations have single, centralized systems to manage their employee related activities.
VIM (Vendor Information Management)
After Customer, this domain has most direct influence on organization commercials. Most of the CDI principles and benefits also apply on VIM. VIM enables you to
- Maintain blacklisting of vendors,
- Negotiate better deals as Vendor Data consolidation gives the true relationship value
- Consolidated Billing and Payments
- Standard contractual and legal agreements etc..
NOTE- Sometimes Vendor, Employee and Sales Agents are also included in the scope of CDI.
Parts Information Management
This is one level down from the Product Information Management. This domain is not much different from the way you do Product Information Management. Parts information Management domain is very closely linked with the production process.
Others
There can be as many MDM domains as there are Masters. Here are some more examples:
- Location Information Management - An organization can have hundreds of thousands of locations (like Sales outlets, ATM machines, Kiosks...).
- Distributor Information Management- This includes information on Agents, sub-agents, brokers etc..
|
