Master Data Management: A Matter of When and How Rather than If

Originally published 2 June 2010

Master data management (MDM) and information governance are going mainstream. Nowadays, companies are becoming increasingly aware of the critical importance of their information assets and the need for efficient data governance. 2010 has so far signalled that master data management has reached new heights in the maturity curve:

  • All the major software vendors have now chosen their strategy and even those that didn’t have solid foundations in MDM have now established this through acquisitions.
  • Some vendors are transforming the market into a volume business with solutions and business models targeted for simpler use, lower budgets and shortened time prior to seeing benefits.
Everything suggested that this year could really set off the master data management market. But nobody could have predicted what a spectacular impact MDM would have had in the first two months of 2010 for the key software vendors:
  • From as early as 4 January, Oracle kicked off the debate, with the acquisition of Silver Creek Systems, (a provider of data quality and data management solutions).
  • Several days later, Talend launched its own open sourceMDM solution, having purchased Amalto.
  • On 29 January, Informatica formalised its purchase of Siperian, one of the “pure players” of the MDM market. This announcement came as no surprise since the software provider had already been making targeted acquisitions to pave the road toward this market over the previous few months. 
  • On 3 February, IBM responded similarly with the acquisition of Initiate Systems. Despite numerous rumours in the few days prior, this acquisition was still difficult to predict as IBM was already positioned as one of the world leaders in MDM. Note that this acquisition is rumoured to have cost IBM between $300-450 million. 
  • Microsoft launched its latest release, Microsoft SQLServer, with embedded MDM, known as Microsoft Data Services. This revelation was no surprise as it followed the acquisition of the MDM pure player Stratature by Microsoft in 2007.
As exhibited by these major market developments, IBM and Oracle are clearly strengthening their portfolios in the MDM industry, while Informatica, Microsoft and Talend are joining a healthy and well defined market, alongside other large vendors such as Information Builders, SAP, SAS Institute and Tibco. Even in 2009, the MDM market proved to be incredibly healthy with companies such as the pure player Orchestra Networks, which had more than 100% growth for its fiscal year of 2009.

A Market Challenged by Industry Demands

The MDM market is, definitively, in the spotlight. Over the last twenty years, businesses have had a tendency to neglect the data dimension of their information systems. Veterans among us remember that this was not the case in the ‘80s and early ‘90s when data models were so proudly advertised on posters in IT offices everywhere. With the implementation of large commercial off-the-shelf packages (ERP, CRM, etc.), object-oriented methodologies and SOA architecture, data models moved into the background, leaving business processes at the surface as the common language between IT and business.

However, times change and business priorities have dramatically evolved. Take e-commerce as an example. Two or three years ago, operational excellence was what made the difference. Companies who were able to commit themselves to punctual deliveries stood out from the crowd. Nowadays, not only is operational excellence a must rather than a competitive differentiator, but also it has become more and more difficult to monetize: delivery fees are now often taken care of by the retailer. What now differentiates online sales websites is not so much their management processes as their capacity to inform and to advise via personalised recommendations. This enables retailers to offer precise, attractive and detailed product descriptions aligned with client needs, customer opinions, expert advice, etc. Well-known websites like even hand over the retailer role to second-level retailers when they cannot deliver the product themselves. This attributes further importance to the value of information management, as retailers like rely on the ability of a third party to deliver on their promises.

As a result, data as a primary resource is assuming its position as the strategic asset that enterprises need to differentiate themselves. Businesses must develop accordingly to cope with this. It has become a necessity for companies to have an information management system that crosses organisation, processes and underlying process-centric information systems. This also requires a uniform governance policy to control data quality, data sharing and security.

This necessity has been driven by new industry requirements and demands for transparency forcing financial and risk departments to take accountability for their data content and sources. R&D has also been pressured to accelerate deadlines for product referencing from the moment of introduction onto the market (marketing, sales, etc. are all seeking improved and multi-channel customer vision). An example can be found in the life sciences industry and its need to reduce cycles for clinical trials. IT has become the key tool for the necessary standardisation and automation of information exchanges between pharmaceutical companies and physicians, hospitals and regulatory activities across countries. Companies can also benefit from this to optimise financial processes such as fast close. This is not to mention the optimisation of vital enterprise-wide data that all activities rely on to operate (such as shared employee and stakeholder repositories, organisational and chart of accounts hierarchies, real estate and asset-related information, etc.).

Another reason that MDM is currently increasing in momentum is its alignment with the current CIO agenda:
  • To help preserve and further value the existing application legacy, rather than needing to invest in new generation systems to replace previous generations’ systems in place,
  • To federate the multiplicity of information sources, internal or external, structured or non-structured, and
  • As the missing piece of the puzzle to deliver on the SOA vision, in order to ensure that each key component of the system is ready to be sourced “as a service” without having to worry about transcodification of data.

MDM, a Descendant of CRM or ERP?

For the last 5 years, MDM has seen pioneers delivering on their vision, while others were still involved in debates. In the same way as CRM or ERP, MDM value relies on its ability to be applied to company activities. As a result, MDM requires close alignment between business and IT. It impacts established roles and responsibilities and therefore requires rigorous change management and the definition of new processes and workflows. MDM is a journey, rather than a destination and demands dedicated teams to operate on a day-by-day permanent basis, rather than in project mode.

In the same way as ERP and CRM, MDM initially originated from several large projects led by a handful of pioneers. The results obtained were significant enough to convince more and more companies to take the plunge. Following this, MDM revealed itself as more complex than originally believed, mostly for the reasons explained above: MDM is not a silver bullet empowered by innovative technology but is about establishing new processes, organisations and roles. This caused many businesses to be wary of potential restraints arising from its complexities and to believe that the implementation of MDM solutions could result in a loss of time, effort and money: the famous “trough of disillusionment.”

As with ERP and CRM, however, a new wave of projects has now been rolled out. In this new wave, the technologies have been refined by time and better control and come together with best practices, taking on the lessons learned from previous projects. Methodologies have been redesigned to be more iterative so that costs before reaping the first benefits are more reasonable. And, above all, there are a wide variety of hotspots in which to progressively launch an MDM project:
  • Departmental MDM projects, or those limited to certain organisational zones within the company.
  • MDM-focused departments like organisational structures, charting accounts, assets or locations to develop initial “quick wins” before moving onto larger and more complex projects, such as client data integration or product information management.
  • Implementation of analytical MDM to find areas of improvement rather than to run the processes. In this area, MDM provides short-term benefits through improved data quality, conformed dimensions across analysis domains to bring the business intelligence (BI) environment to the enterprise rather than the departmental level, or ability to cross analyse structured and unstructured data based on common master data.
  • Modernisation of data-administration applications and reference table management from legacy applications: many companies, notably banks and insurance, implemented systems in the late ‘70s or ‘80s to manage reference tables that were and still are shared by their mainframe systems. Although these systems are still mandatory to run the business, they are outdated, poorly documented and depend on a few employees to be properly managed and maintained. MDM is a good way to implement new systems and processes, while avoiding the risk of a complete system overhaul.
Time has also helped us to find ways to progressively implement MDM, by adding services one by one, on the way to a well-defined maturity model. Typically, an MDM initiative can be set up through a four-step process:
  1. Assessing data quality can be the first step since it helps to gauge the current situation, providing the measures that will be used to prove progress and identify the return on investment.
  2. The second step could be automatic data correction (when this can be automated). This step is important since it can deliver quick wins, showing that the process is running with concrete results.
  3. Following this, a data governance process must be established and implemented, defining responsibilities and processes for both certification and data protection. This step is mainly about change management, communication and process definition.
  4. MDM presents itself as the solution to empower the previously defined processes. It provides the foundations to operate these processes on a repetitive, efficient and consistent basis.
In the ERP and CRM industries, experience has so far proven that although the value of these solutions is to achieve holistic integration and synchronicity across an enterprise, a hard-liner implementation of these concepts can clash with the practical realities: few companies have one unique completely integrated authority on these solutions. It is presumed that MDM will follow the same pattern.


According to Wikipedia, master data management is a set of processes and tools that consistently defines and manages the non-transactional data entities of an organisation (which may include reference data) as the objective of providing processes for collecting, aggregating, matching, consolidating, quality-assuring, persisting and distributing such data throughout an organisation to ensure consistency and control in the ongoing maintenance and application use of this information.

SOURCE: Master Data Management: A Matter of When and How Rather than If

  • Jean-Michel FrancoJean-Michel Franco
    As Innovation and Solutions Director for Business & Decision, Jean-Michel Franco designs and markets value propositions on an international level. Franco has dedicated his career to developing and broadening the adoption of innovative technologies in companies. He started out at EDS by creating and developing a new business intelligence (BI) practice. Franco then joined SAP EMEA to develop the business in the areas of BI and ERP solutions and later became Director of Marketing Solutions in France and North Africa. He may be contacted by email at

Recent articles by Jean-Michel Franco



Want to post a comment? Login or become a member today!

Be the first to comment!