[This article of mine was first published in ECCMA's bi-monthly news letter]Master data management is an absolute key area to the success of eProcurement and inventory optimization initiatives. Traditionally, Buyers have relied on Suppliers to provide product information. There is also a casual attitude about the inventory data that it is after all a simple medium of identification of parts and assets.
In fact, inventory data management is much more complex than it appears at first glance. It is therefore important to realize the challenges of managing the inventory data and put in place a robust and long term operational strategy to face it.
First, let us understand why it is important to manage Master data?
- Master data is the life-blood of any successful eProcurement transaction
- It is a key element in the decision of planning and budgetingIt is the fuel for accurate spend reporting (Eg: Material aggregation by commodity groups)
- Good Master data helps to achieve regulatory complianceIt assures increased uptime and reduction in stock-out situations
- All of the above leads to faster executions, quickened ROI and increased customer satisfaction
Now let's also try to understand the common woes of master data management. Perhaps the major problem lies in the fact that an organization's control over the content is often poor. This could be due to various reasons, like;
- Fragmented source systems and/or fragile integration between them
- Unclear data ownership
- Poor inventory control proceduresInadequate focus from top management
- Differing usage of semantics (Ex: ID / Inside dia, In/Inch/" etc)
Consequence of the above could prove to be very costly for the enterprises.
The most common early 'symptoms' of this problem are;
- Duplicate and redundant data found in procurement systems
- Increased reliance on the "outside the system" for data and reporting
- Increase in inventory levels
- Noticeable drop in order fulfillment
- Deviations from standard procedures (workarounds) to meet the business needs
If these 'symptoms' are not addressed with a definitive strategy, it would soon turn into a enterprise-wide 'epidemic', severely limiting the organization's ability to take advantage of the IT infrastructure.
Taking control : Strategy for effective master data governanceEffective master data governance is not a stand alone subject confined to a particular function. It is about putting in place a corporate-wide policy towards Supplier enablement, technology, workflows, procedures, standards, roles and responsibilities and continuous monitoring.
While each of these deserves a discussion in detail, this article's scope is however limited to discuss only the data issues.
Best practice approach for master data governance consists of these four steps, each of which is a small project in itself.
- Data cleansing and harmonization exercise
- Supplier enablementWarehouse organization
- Process for ongoing maintenance of Master Data
Out of the above, let's look at step #1 in more detail, the most important and perhaps the most challenging among the four.
Data cleansing and harmonization : A well planned data cleansing exercise, either done in-house or through a professional service provider should address the below needs:
- Item identification
- Classification coding
- Usage of parametric data (attributes)
- Application of standards
- Enrichment
1. Item identification
A combination of part number and an accompanied brief description is often necessary to unambiguously identify a product. Therefore a data cleansing exercise must focus on providing accurate part numbers and ensuring completeness of product description. The short descriptions must be "normalized" (removal of excessive abbreviations) and "rationalized" (logical ordering of description by attribute)
2. Classification coding
It is highly recommended that Master data be classified to a hierarchical classification standard. UNSPSC (Universal Standard products and services code) is by far the most popular and natural choice. UNSPSC offers a single coding convention that covers all products/services (well, almost all) and provides the broadest collection of businesses, industries and commodities available today. The advantages of UNSPSC are;
- Quick and easy 'drill down' search of product information
- Enables spend analysis by commodity grouping through "Rolling up" the hierarchies.
For more information about UNSPSC, see http://www.unspsc.org
3. Usage of parametric data
Parametric data (attributed data) is a necessity rather than a luxury for plant equipment and spare identification. They help in providing specifications to the user in a clear and concise format.
Though there are few proprietary attribute standards available for a fee, by far the most popular, universal and recommended 'open' attribution standard is ECCMA's eOTD (ECCMA Open Technical Dictionary). eOTD is not merely an attribute standard, it provides a complete framework of defining the product including a standard class name, list of properties, standardized representation of values, UOMs and a cross reference to various other product classification standards.
For more information about eOTD, see http://www.eccma.org
4. Application of standards
Application of standards leaves little space of ambiguity. Standards can be applied for a number of data elements at different levels within the Master data. Following data elements pertinent to eProcurement have a defined set of standard denotations.
- Currency codes
- Unit of Measure (UOM) codes
- Location codes (locales)
- Language codes
5. Enrichment
Enrichment is an optional exercise that if undertaken would add depth to the content. The more information is provided to the user, the better will be the purchasing experience. Adding a Long description, Picture, Drawing, MSDS, Keywords etc would enhance the user experience.
Conclusion
Studies conducted by reputed research organizations on data quality have shown that very low % of data in the inventory systems of various businesses are in usable and transactable form. They also point out the facts that;
- Technology by itself doesn't fix the data quality issues
- Organization's overall e-business strategy is incomplete if it doesn't address content management issues
- Time and effort involved in fixing content issues should not be under estimated
- Cost of cleansing the content is a fraction of opportunity cost!
- Perhaps the most important lesson learnt by many organizations is that trying to manage the content manually in-house absorbs time and resources.
- Its better to let the professionals to manage them.