Guest blog post by Guillaume Duval, Senior MDM Consultant, CGI
How is master data management evolving to help organizations unify data governance, master data management, cloud, big data and social networks? What are the key trends in MDM? Guillaume Duval, Senior MDM Consultant from CGI, sought answers to these question and more at the recent MDM/DG Summit in New York. In this guest blog post he shares his thoughts and observations about the MDM market and market trends.
Over the past 10 years, Master Data Management and Data Governance has evolved from a set of IT-centric back-office tools to business-facing solutions able to facilitate distributed data management processes across the organization. With a decade of experience, maturing data governance techniques and new technology capabilities, the industry may be on the verge of another evolution: MDM 3.0.
Although the latest generation of MDM is emerging, not all vendors are currently delivering the most evolved MDM. To help you distinguish the different generations of MDM, we offer you this digest of developed from content presented at the MDM/DG Summit.
MDM 1.0: MDM designed for IT and purpose-built for a single domain
Analysts describe MDM 1.0 as the first generation of MDM initiatives that brought a “single version of the truth for a specific domain” between the information systems. MDM 1.0 solutions are characterized by passive data governance. Data stewardship is only invoked after-the-fact.
These first generation of solutions were all purpose-built, designed to manage either customer data integration (CDI) or product information master (PIM). PIM solutions focused on managing product, supplier and pricing information through different catalogs structures. CDI solutions focused on managing the contact information for people: retail customers, citizens or patients.
The domain-specific design of MDM 1.0 solutions are ideal for narrow MDM programs that do not benefit from information in other domains. Some analysts point out that the limited scope make it easier to sell MDM 1.0 to the business as a tactical, targeted program. This is due to the sensitivity, visibility and high presence of customer and product data throughout in information systems. However, there was one caveat. “Tactical wins” can become “Strategic failures” if the organization needs to expand from these domain-specific solutions to a more broadly-scoped next Generation MDM.
MDM 2.0: MDM adds data governance and business user support
MDM 2.0 describes the initial convergence of MDM, data governance (DG) and business user support. One element that separates MDM 2.0 solutions from MDM 1.0 solutions is how they are delivered.
MDM 2.0 vendors offer MDM, DG and business user support all in one solution. 2.0 vendors can get to a single solution either through marketing bundles, where the components are distinct elements but fall under a single price/license, or integrated packages, where all the components are part of a single integrated solution. While there is not universal agreement about the superiority of marketing bundles vs. integrated packages—some have pointed out that integrated packages tend to be more cost effective to implement and maintain. In contrast, MDM 1.0 vendors separately offer, price and license their components for customer data integration (CDI), product information management (PIM), reference data management (RDM) and Data Governance (DG).
The biggest gap between MDM 1.0 and 2.0 is that MDM 1.0 delivers IT value rather than business value. In MDM 2.0 workflows become true, distributed data governance processes involving business users in authoring, review and approval. This is in contrast to the after-the-fact, or ex-post, “break-fix” workflows found in MDM 1.0. Additionally, MDM 2.0 data silos are much more integrated, the scope of data expands to include data passed from internal systems to business partners, and the enterprise starts considering unstructured data from the domain.
With MDM 2.0, it is no question that Data Governance becomes more and more automated. Data Governance, based on metrics, following an upstream data policy replaces manual administrative processes with role-based and real-time workflows. Other topics emerged in our discussions around MDM 2.0, notably Multi-entity MDM (parties hold multiple roles, your employees can also be your customers) and reference data management.
While reference data management, with its focus on foundational master data, seems to offer less immediate value to the business, several analysts, including Aaron Zornes, pointed out without a reliable foundation of reference data building a sustainable data governance competency in the organization becomes even more challenging.
MDM 3.0 Multidomain relationships, Social Networks, Cloud, and Big-Data
The consensus is that the next generation of MDM, 3.0, will need to address multidomain relationships, social networks, cloud integration, big data & Mobile.
While social, cloud and big data are all important, one of the key findings from the 2013 edition of the MDM Institute’s “Reality Check” survey was that large enterprises are actively seeking technologies, today, that can support the management of multidomain relationships and hierarchies between different entities/parties. The MDM institute also found that these large corporations do not think that existing MDM 1.0 platforms and their derivatives possess the ability to manage relationships capably to support standard and alternative hierarchy management. Indeed, based on our own experience at CGI, the hierarchy management is an important feature for MDM. The ability for an MDM solution to manage the relationships that underpin a hierarchy is a considerable asset.
Relationship management is also important for Big Data in analytical use cases. As large enterprises start to delve into Big Data projects, they will find that careful management of enterprise dimension values and relationships over-time periods will be of vital importance to business intelligence. MDM 3.0 with its ability to govern multidomain relationships across domains and time will be useful for the enterprise dimension management.
Nearly everyone agreed that today social networks are of limited use for organizations that have B2B business models. However, big “social” data is useful for organizations focused on individual consumers. Interestingly enough this form of big data is breathing new life into the purpose-built components of the MDM 1.0 vendors. High quality identity resolution, commonly used to create a ‘single view of customer’ from internal systems data, can be used by organizations that wish to tap, trace and market to individuals across social networks. The challenge is that given the mixed environment (on-premise and cloud) these enhanced MDM 1.0 implementations may require persisted, registry & hybrid architectures that some of the solutions currently do not possess.
Concerning the cloud, which is one of the more interesting trends for MDM 3.0, there is continuing reticence to host customer & other proprietary master data in cloud. This is especially true in countries where privacy regulations have restricted transnational movement of personally identifying information without consent.
That said, some have found that economic and time-to-market concerns and are over-riding standard IT objections against cloud. While it remains to be seen, we may see a world in which Cloud MDM manages non-core master data, such as public reference data, while on-premise systems master proprietary and customer information (including personally identifying information). Some individual said that the deployment method is a red-herring as it does not change the requirement for MDM (3rdgeneration or not) to supply SOA/shared data services that support an organization’s entire application architecture.
Finally, MDM solutions will integrate data governance-ready frameworks. Data governance solutions will pass from active to proactive management by enforcing data stewardship and control in real-time with distributed workflow capability for business users & IT to actively control their shared data across lines of control.
A good example of how technology vendors can satisfy the requirements for MDM 3.0 comes from one of our partners, Orchestra Networks. In the latest release of their multidomain master and reference data management software, EBX version 5.4.0, Orchestra Networks added metadata management to its already substantial data governance features. EBX5 also includes a wide array of capabilities for hierarchy management and data modeling that satisfy the needs of large organizations that need to manage of multidomain relationships and hierarchies. EBX5’s ability to manage relationships between domains, and even time, is important for both enterprise dimension management, and consequently big data analytics. Finally, unlike many MDM platforms, EBX5 delivers all of its capabilities in a single integrated package instead of marketing bundle, or as stand alone, non-integrated applications.
MDM 4.0 and beyond
While MDM 4.0 is a long way off, analysts, product marketers and consultants did not shy away from prognostications. Some speculate that MDM 4.0 will become MDM as a service, or “pervasive MDM” as MDM vendors and SaaS vendors adapt their technologies to be used in a more integrated way.
Other assert that MDM 4.0 will address, completely, the big “social” data issues, to finally unify marketing and customer data management. This may be where companies that have a B2C business model finally receive some benefits from Social MDM, as sentiment analysis on tweets and status updates translate into more informative net-promoter scores that enrich customer relationship data. It may even opens the means to manage and market to the same party as they take on different roles and functions or cross geographic boundaries.
Many think that 4th Generation MDM solutions, fusing hub and registry styles, could begin to incorporate unstructured information. Could there be a conceivable tie up between MDM and document management in our future?
The common belief is that at this stage, Multi-MDM and Reference Data Management will be inevitable. Actually the business value of multi-MDM allows enterprises to be able to:
- Identify & provide differentiated service to its most valuable customers, via their relationships (households, hierarchies)
- Introduce new products & product bundles more quickly across more channels, to reduce the cost of New Product Introduction (NPI)
Provide improved enterprise-wide transparency across customers, distributors, suppliers, and products to better support regulatory compliance processes
Recapping the key points
To summarize, here are the main MDM trends for today and in the coming year 2014:
- MDM is a mature market
- MDM market is now mature and popular
- Clients are convinced of its interest and advantage
- Transition from domain-specific to multidomain solutions
- Cloud MDM to Pervasive MDM (SOA enabled)
- Social MDM is the management of big “Social” data to help market and manage consumers
- Evolution from Active to Proactive Data governance
- Sophisticated standard and alternate hierarchy management is becoming a must
- Measurements / indicators
- Security management
- Data confidentiality / Data masking
- Data segregation due to Organization roles, profiles and permissions
- Workflow to help data governance with regulation and conformity
- International MDM: MDM solution evolve to ERP model with solution based on core model and roll out locally