Master data management (MDM) has become a priority for large organizations.
Business realities such as global outsourcing, the need for differentiated customer service, better risk management, and more efficient internal operations are forcing organizations to get better insights and value from their information assets. It strategies to support these initiatives, such as service-oriented architecture (SOA), do not realize their full return on investment when the underlying data they rely on
is inaccurate, inconsistent, and often not rich enough to provide business value.An MDM solution enables businesses to align their enterprise master data assets (product, customer, vendor, etc.) across multiple systems and departments and with trading partners. It also enables organizations to support the necessary processes, policies, and procedures to ensure that accurate and consistent
information is disseminated across the organization to the transactional systems and decision-makers that rely on this information for their day-to-day operations.
Businesses today must constantly roll-out new value-added services to stay competitive.Retail and manufacturing firms are struggling to meet mandates around collaborative planning, forecasting, and replenishment (CPFR), global data synchronization (GDS), and radio frequency identification (RFID). With mass consolidation in financial services, firms are looking to gain a single customer view for up-sell/cross-sell campaigns. telecom companies, faced with increasing churn rates and lower switching costs, are focusing on customer service initiatives and better service provisioning. trading desks across energy, commodities, and securities are trying to better profile and track counterparty risk. Nearly every
industry can benefit from better customer service, more efficient procurement, consolidated billing, and more visibility into operations across departments. To support such initiatives It departments are building strategic composite applications on top of a service-oriented architecture (SOA) backbone. In order for these service-oriented business applications (soBA), or composite applications, to meet the needs of the business community, It must achieve some level ofinformation alignment so that different systems being accessed by a business service have consistent definitions and values for critical master data such as product names. yet most companies do not have accurate and consistent
information across people and systems, either internally or across their value chain. This lowers the return on their It investments.
A master data management (MDM) solution enables businesses to manage and align their enterprise master data assets (product, customer, vendor, etc.) and build and support the necessary processes, policies, and procedures to ensure that clean data stays clean. master data should not be confused with transactional data. for example, product master data refers to product attributes such as product name, description, dimensions, and other such attributes. Individual order information such as po numbers, order quantity, etc. are examples of transactional data. A custom attribute such as credit score, which might be computed using transactional data and updated more frequently than a static
attribute such as customer name, is still considered and managed as master data, as it is a descriptor of who that customer is, similar to ‘customer address’ or ‘income bracket’. managing such a rich and complex array of ever-changing and interdependent attributes across multiple data domains on an ongoing basis and then ensuring that this data filters down across the multitude of transactional
systems and decision-makers that depend on it for their day-to-day operations is the power of an MDM solution.
Aligning master data internally across multiple instances of enterprise resource planning (ERP), customer relationship management (CRM), and supplier relationship management (SRM) systems, to name a few, is only a first step, albeit an important step. Just as important is making sure that information – such as product names and attributes or customer data – is consistent across the value chain so that companies can reap the benefits of stronger value chain collaboration such as smaller inventory buffers in a supply chain, better customer service for large customers, and reduced order-to-cash cycles through more efficient purchasing and invoicing.
Master data management refers to the generic problem of managing enterprise master data assets and empowering downstream people, systems, and partners with the information they need. different organizations may choose to focus on a different area or domain of master data. Nearly every company can benefit from more accurate information on their products, customers, and vendors across
industries. Common challenges that MDM addresses include:
- A large number of products, parts, or assets, each with multiple attributes that need to be managed and enriched by different departments
- Need to gain more insights into customers for both new revenue opportunities as well as better customer service across multiple customer interaction points and channels
- Need to harmonize inconsistencies across systems, business units, and geographies
- Need to execute business processes such as new client introduction that cut across information silos and organizational boundaries – geographic and functional
- Need to receive client and security information from external feeds such as reuters and Bloomberg for trade settlement
- Need to collaborate with other businesses in the value chain to achieve efficiency
- Partner and regulatory compliance mandates requiring the management of new types of information and the maintenance of data lineage trails
- Need to measure metrics that require aggregation across multiple systems such as counterparty risk across all trading assets
- Need to gain efficiencies in procurement across multiple products and vendors
Most companies will choose to tackle one data domain first and then use that experience to expand into other domains. many industries will have data domains that are uniquely valuable to them. For example retail chains may want to first focus on store information. An auto manufacturer may want to manage dealer information. healthcare companies will have unique challenges around patient information. While these different data domains may appear very different to a business analyst, from a technology standpoint they pose similar challenges.
Comprehensive Information Management: Managing the data model and attribute information, validation and transformation rules, versioning and diffanalysis, roles-based access and ownership control, complex data relationships including management across data domains, classifications,
contextual validation rules, etc.
Process Management: Managing the processes and procedures around introducing new data or editing existing data such as launching a new product or updating a customer address.
Integration: Synchronizing in real-time or batch the relevant subset of master information with transactional systems such as ERP and trading partners either directly or through exchanges and data pools such as 1synch.
MDM DRIVES ROI FOR SOA INVESTMENTS
An MDM solution provides the necessary alignment of master data across multiple back-end systems so that business services and composite applications within an SOA have accurate, consistent, and timely information. All the hype and attention in SOA deployments has gone into web service creation, deployment, and management standards and technologies. However, if data is inconsistent across
applications it will be increasingly difficult if not prohibitive to build composite applications that cut across multiple systems and departments. For example a composite application in a large multi-channel financial service institution that calculates a customer’s global credit risk will only work if that customer is described in a consistent manner across retail banking, brokerage, mortgage, and credit card systems. In a retail environment, a composite application that gets a customer’s order history from a data warehouse and recommends a related product will require consistent product and customer information across all the relevant systems. On a smaller scale, even business services to update an address or provision a service require semantic consistency of master data across CRM, billing, and product systems. Creating a semantic integration layer to harmonizeand reconcile master data across the enterprise yields accurate and consistent information that helps SOA investments realize their full ROI.
HARMONIZED – TIBCO’S APPROACH
Given the pitfalls in storing master data in a dominant application or having a completely federated approach, TIBCO believes that the harmonized model for MDM yields the most value for customers. TIBCO collaborative Information manager software has the following attributes that make it unique in the marketplace.
Harmonized:TIBCO builds and maintains a central master data repository that serves as either the system of record or reference and can be a source of authoring of information or designate authoring to transactional systems. As the system of reference it keeps track of ownership at the attribute level and applies precedence rules to ensure survivorship is accurately handled when importing new or changed data. The repository then synchronizes with multiple downstream transactional systems and trading partners that depend on that data. the master data repository can also serve master data as services to requesting composite processes over a Web services API. The repository is populated by aggregating,cleaning, transforming, and enriching data from internal and external sources and maintained by the process-centric capabilities of the application.
Process-Centric Approach:TIBCO takes a process-centric approach to information management by providing customizable processes to introduce and manage data, such as a “new product introduction” process in a manufacturing company or a “new vendor introduction” process in telecom. If data is cleansed and aggregated, such as in data warehousing projects but the necessary business process automation and data governance policies and controls to introduce and update new data are not put in place, the value of the solution will be fleeting and over time the original problem of data inaccuracies and inconsistencies will return.
Collaborative:The process-centric nature of the application allows it to be extremely collaborative. Different attributes of data will always be owned and managed by different business users. For example, product brand data will be managed by marketing, packaging data by logistics, and product specs by R&D. TIBCO CIM facilitates cross-department and cross-geography collaboration to manage, update, and enrich master data.
Multi-Domain:TIBCO provides a single application to manage data across multiple strategic data domains, be it customer, product, vendor, etc. It does so through an extensible configurable data model and customizable information management processes to meet the unique needs of different types of
information. managing multiple domains on a single platform leads to a low TCO and allows customers to scale their solution as their needs expand. In addition, customers can also manage relationships between data domains such as between products and vendors, or securities and clients.
SOA:TIBCO has over 20 years of experience in distributed computing and has pioneered publish-subscribe technologies. TIBCO leverages these strengths to provide a highly distributed solution that can scale to meet the demands of very large deployments. The TIBCO CIM repository is synchronized via both batch and real-time technologies with internal and partner systems. Through features such as the
distributed cache, Web services API, support for multi-threaded asynchronous processing, event-based architecture, and embedded JMS client, TIBCO CIM is the leading solution for customers with SOA initiatives or looking to support SOA principals for the long term.
Peer-to-Peer:TIBCOCIM supports peer-to-peer synchronization with other TIBCOCIM instances. Such an architecture might be necessary in a global organization where regional or business unit differences are so great that it makes sense to keep a local master data repository in each region or department and synchronize it with corporate headquarters and other regions. This is common in multi-bannerretailers, or large conglomerates looking to provide regional autonomy with centralized visibility and control.
A Complete MDM Solution
For an MDM solution to be successful it has to be tightly integrated in a bi-directional manner with the underlying It landscape as well as trading partner systems, which are out of your control. TIBCO CIM supports multi-channel bi-directional state-full synchronization with internal systems, trading partners, and industry exchanges such as 1synch.
The solution must be flexible to account for the fact that the link with every system will be governed by its own set of data transformation and validation rules, synchronization frequency (batch, near real-time, real-time), and integration transport (ftp, point-to-point JMS, web services, message bus, As/2).
During the on-ramping of the solution, data needs to be profiled, cleansed, aggregated, and loaded in bulk with validation rules into the repository. After deployment, the entire solution, including the core TIBCO CIM application and all data feeds going in and out of it, needs to be monitored in a mission-critical manner.
Finally, ongoing data management processes and key metrics have to be tracked for effectiveness for continuous process improvement.
TIBCO provides a complete MDM solution with a high-powered business integration platform encompassing enterprise application integration (EAI), business-to-business (B2B), enterprise service bus (ESB), and extraction, transformation and loading (ETL) capabilities as well as necessary monitoring
and management. All TIBCO products, while designed to integrate well with each other, are open and standards-based. hence, the application can work with whatever integration platforms customers already have in-house. While TIBCO provides all the supporting components mentioned above, no other TIBCO products are required other than TIBCOCIM for MDM. Customer who wish to leverage their existing middleware infrastructure can do so. TIBCOCIM can be deployed as a stand-alone MDM application.
Synchronizing information across the enterprise enables organizations to achieve the following business benefits:
- Accelerate time-to market through process automation and information sharing for new product introduction processes and supply chain collaboration.
- Increase revenues with cross-sell and up-sell opportunities through single customer view.
- More efficient trade settlement and reconciliation for securities firms.
- Increase customer satisfaction through a multichannel unified customer experience.
- Reduce shipment inaccuracies and invoice errors with trading partners.
- Be able to analyze and optimize procurement through a single view of vendor.
- Get a global view of risk across counterparties.
- Increase ROI on SOA and BPM initiatives by providing business processes and composite applications with accurate and consistent information