Pushing Higher

Master data management is a strategic competitive business strategy. Is your business looking for new ways to drive down costs or enable better regulatory compliance? What about enabling growth as the economic upturn comes or when mergers and acquisitions are on the horizon?

Whether increasing revenue, improving risk management and compliance, optimizing operational efficiencies or strategically differentiating an organization from its competitors, MDM is widely recognized at the executive level as a compelling proposition-usually manifested in a business initiative called "single view of the customer" or a similar name.

MDM provides a trusted, consistent view of key information assets across the enterprise-ranging from customers, products and suppliers to locations and more. In large corporations, MDM is becoming a business transformation strategy as the cornerstone of every critical business process and business decision.

Both master data management and data governance programs are breaking new ground with business value justification for large-scale integration projects that blend brick and mortar with e-commerce programs. MDM takes the notions of integration and transformation to new heights. By treating data as a corporate asset, an enterprise MDM program proposes to integrate front office and back office as well as (often competitive) business units within the enterprise.

Why is governance of master data so important? MDM solutions for a unified customer/citizen/supplier view and similar data integration initiatives range from do-it-yourself initiatives to commercial off-the-shelf solutions such as data hubs and registries. Both variations require data governance solutions to navigate the physics and politics of large-scale data management. Manual data governance is error prone, time-consuming and unable to ensure compliance or measure business impact. However, the addition of passive data governance (such as data steward consoles) to MDM does not fully leverage data quality, because business processes themselves must be enabled, synchronized (via business process management) and aligned with the data.

What the market requires for enterprise-strength MDM is "proactive" upstream data governance-ideally integrated with an MDM platform so that the documented policies become operational policies.

Over the past six years as conference chairman for the MDM & Data Governance Summit, I have observed hundreds of MDM implementations in almost every industry around the world ranging from very large, highly heterogeneous distributed enterprises to midsize, mostly homogeneous centralized/local enterprises. In this same time frame, more than 5,000 IT professionals have attended our workshops and tutorials in London, Frankfurt, Madrid, Moscow, New York City, San Francisco, Singapore, Sydney, Tokyo and Toronto. In turn, we have distilled our findings into candid insight and experience-based guidance for IT professionals embarking upon their MDM and master data governance programs.

Moreover, during the first quarter of 2011, analysts at the MDM Institute reviewed more than 750 MDM and data governance case studies as part of the process to arrive at the 10 strategic planning assumptions reviewed in this summary report. The typical audience for such guidelines includes CTOs, CIOs, enterprise data and solution architects, vendor product management and product marketing, and investors.

 

MDM Roadmap for 2011-2012

Part of the deliverables for our client advisory council is an annual set of milestones and a roadmap to help Global 5000 enterprises focus efforts for their own MDM projects. Ten to 12 milestones are explored, refined and published via our MDM Alert research newsletter. This set of strategic planning assumptions presents an experience-based view of the key trends and issues facing IT organizations across areas of master data governance, customer data integration, product information management and reference data management. Thus the 2011-12 MDM roadmap helps Global 5000 enterprises (and IT vendors selling in this space) utilize these strategic planning assumptions to focus their own roadmaps on large-scale and mission-critical MDM projects.

Market maturation (consolidation and diversification). During 2011-12 as MDM platforms move from third generation (service-oriented architecture, single domain focus) to fourth generation (BPM-enabled, data as a service), the mega vendors (IBM, Informatica, Oracle, SAP) will dominate while specialty solutions proliferate by industry, use case and geography. Application package vendors will have rearchitected such that their 2011-12 solutions are MDM-innate rather than MDM-enabled.

Master data governance. Data governance of master data will remain problematic through 2011-12 as marketed solutions lack systemic rigor and E2E lifecycle support. Even more desirable is proactive integrated governance, where an asset-focused methodology is applied, rather than passive in which a mismatch of applying project-oriented methodology is created. During 2012-13, frameworks addressing the "community" aspect of shared asset development (e.g., wikis) for global corporate business vocabulary will arrive.

Policy hubs. During 2011, MDM solutions providers and BPM solution providers will increasingly collide in the market as the latter acquire or build out BPM-centric MDM. Through 2012, however, BPM-centric MDM will suffer from BPM's traditional focus modeling but not executing MDM rules. By 2013, all mega MDM and BPM vendors will have overcome this dogmatic bias. Moreover, enterprise BPM needs to execute within governance and vice versa (i.e., be able to execute MDM workflows within BPM) by sharing state, process and events. Lastly, enterprises apply different business processes for CDI versus PIM that the fourth generation master data solutions need to heed. To further complicate matters, there will be an ongoing overlap between MDM solution providers and application package vendors. It will be interesting to see how far up (or down) the stack MDM vendors and app vendors will go.

MDM convergence (multi-entity/domain). During 2011, corporate MDM platform evaluation teams will assume and insist that all MDM software platforms targeted on enterprise-level deployment (or even playing a major role in mission critical systems) fully support both PARTY and PRODUCT entity types. In my experience, RFIs/RFPs assume that "MDM" means "multidomain MDM." By 2012, large enterprises will also increasingly mandate that reference data management be part of the MDM platform native entity types. Throughout the next two years, operational CDI hub vendors will add light PIM capabilities, and PIM vendors will add B2C PARTY capabilities.

Universal/pervasive MDM (MDM as a service). During 2011, application package providers will deploy en masse the next generation of MDM-innate (as opposed to MDM-aware) applications; concurrently software as a service (SaaS/cloud) vendors will struggle to provide integrated/native MDM. Also during 2011-12, mega vendors Microsoft and Oracle will essentially give away departmental MDM - via Microsoft Master Data Services and Oracle Data Relationship Management, respectively. By 2012, SaaS providers will have finessed this issue via strategic partnerships and investments in MDM solutions. Moreover, the market for MDM-enabled applications will continue to exceed that for standalone MDM software.

Budgets/skills. During 2011, the number of IT professionals trained in a specific MDM solution will increase 200% over the year prior, but IT organizations and consultancies will struggle to recruit and retain MDM veterans who have successfully had a major role in an MDM deployment. In the first year of deployment, enterprises will continue to spend on average three to four times as much on services as on MDM software. Both IT organizations and consultancies will be challenged to retain key employees. By 2012-13, the supply of MDM-experienced consultants will catch up with the pent-up demand of prior years and rates will fall proportionately.

Architecture and data models. Through 2011-12, registry-style MDM solutions will find favor in industries where the data is legally or physically too difficult to consolidate into a physical hub (especially government, U.S. health care). Additionally, mega MDM vendors will apply the powerful hierarchy management capabilities of native registry solutions to integrate both legacy MDM hubs and enterprise content management. Through 2013-14, the mega vendors will adapt the matching algorithms of their registry MDM hubs to bolster the performance and accuracy of their operational/transactional hubs with matching throughput.

Identity resolution. During 2011-12, the requirement for accurate and scalable matching capabilities (identity resolution) will continue to frustrate certain vendors as others monopolize such capabilities for their own MDM platforms. Flexible hierarchy management for both consumer households (demographics) and business entities (firmographics) will look to trusted third-party data sources not just for name and address cleansing but also maintenance of hierarchy relationships-despite poor quality data outside North America and Europe for this purpose. By 2013-14, enterprise search and related semantic technologies will help make MDM more accessible.

Big data/in memory. During 2011-12, the performance of all major aspects of base MDM functionality will benefit from the performance-enhancing capabilities of big memory configurations, from batch loading of MDM hubs to identity resolution to operational updates. Very large enterprises will be looking for real-time MDM flows and scaling of MDM solutions via the elasticity of cloud-based solutions, in-memory cache databases, i.e., the next-generation of ETL/MDM.

Business-critical MDM. During 2011, dogmatic spats regarding analytical versus operational versus collaborative MDM use cases will blur as each of these become business-critical MDM, which in turn demand zero downtime. By 2012-13, MDM platforms will provide in-situ capabilities to change the data model and business rules without taking the MDM services offline. This will impact the way master data services are syndicated and delivered regardless of regular ablutions such as software upgrades.

MDM is quickly broadening its attractiveness as a key enabler of both strategic business initiatives and tactical P&L initiatives. As of 2011, MDM is clearly "for the masses" and business-critical for large enterprises.

Seasoned MDM and data governance professionals acknowledge that successful MDM requires a significant upfront data governance investment, and data governance as a discrete discipline benefits tremendously from the application of MDM software.

These two initiatives are co-dependent in very many ways, yet the market has been late to bring the two together to optimize the people, process and technology. Proactive, integrated data governance will remain problematic during 2011, however, as software vendors struggle to rationalize MDM and BPM.

MDM and data governance are codependent/interdependent. Invest upfront in data governance for MDM sustainability and return on investment. Go early, go governance.

Aaron Zornes is the founder and chief research officer of The MDM Institute. This article was first published in the May/June 2011 issue of Information Management.

 

 

TERMINOLOGY

As convoluted as MDM may seem, there is universal terminology technologists follow to keep it as simple as possible. Following is a list of terms and their standardized definitions.

Data Governance Formal orchestration of people, process and technology to enable an organization to leverage data across lines of business and IT systems

Active Data Governance Metrics-enabled, upstream data policy enablement; replaces manual data admin processes with role-based, real-time subject matter expert empowerment

ProactiveDataGovernance Metrics-driven, crowd-sourced capability for business users and IT to control their shared data across lines of business and IT systems

Passive Data Governance Data steward consoles and other reactive data management capabilities focused on after-the-fact data compliance; often patch-like and not integrated with MDM

 

A GLOBAL VIEW OF MDM

Compliance and single customer view are universal business drivers. Product master data is very well established in Europe, customer master data less so. Enterprises are moving beyond CDI and PIM to include reference master data. Enterprises have stopped calling out multidomain and now assume MDM includes multiple domains. North American enterprises rely more upon mega consultancies than European counterparts. Source: Survey findings presented at MDM & Data Governance Summit Europe-April 2011 in London

 

3 MDM KEYS FOR INSURERS

Master Data Management (MDM) is the best recourse to unlock siloed data, often associated with multiple applications for each single line of business, Joshua Schwartz, director at PwC's Diamond Advisory Services, told INN blogger Joe McKendrick. "For many insurance processes, especially first notice of loss, every second counts," Schwartz said. "Master data can really help in searching and retrieving existing customer records, or pulling in associated claimants to a party among people, companies, vendors or third parties, which we already know saves time up front. This accelerates the processes for adjudicating and processing claims. The time spent by your claims handler or servicing agent to continually repeat and input data not only creates data quality problems downstream, but takes time when servicing customers." Schwartz provided three key elements that should be part of an MDM approach:

1. Address organizational issues by identifying an owner for master data assets and for each domain across business functions; and develop a data stewardship program to enforce consistency and reusability for data delivery and use.

2. Create common and shared processes for defining, administering and governing master data by establishing a formal governance process and decision-making guidelines for master data.

3. Employ technology to establish a single version of master data containing uniquely identified records from various business and information domains. Technology should include an integration platform to support the movement of master data between operational systems and across external entities. Technology also needs to enable the establishment of business rules that define the "trustworthyness" of data sources, and the attributes to develop a golden view of data.

For reprint and licensing requests for this article, click here.
Analytics Data and information management
MORE FROM DIGITAL INSURANCE