BT

Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ

Topics

Choose your language

InfoQ Homepage Articles Improving Data Management with the DMM

Improving Data Management with the DMM

The CMMI Institute has launched the Data Management Maturity (DMM)SM model. It can be used to improve data management, helping organizations to bridge the gap between business and IT.

Using the DMM, organizations can evaluate and improve their data management capabilities. The model leverages the principles, structure, and proven approach of the Capability Maturity Model Integration (CMMI).

InfoQ interviewed Melanie Mecca, Program Manager of the Data Management Maturity product portfolio at CMMI Institute, about why the Institute developed this new model, how organizations can use the DMM to improve data management, combining the DMM with agile adoption in enterprises, and the relationship between the DMM and the CMMI and People CMM.

InfoQ: Can you explain the Data Management Maturity model to the InfoQ readers?

Melanie: The DMMSM model is a unique comprehensive reference model for fundamental data management disciplines. It provides organizations with a standard set of best practices to assess capabilities, strengthen the data management program, and develop a custom roadmap for improvements which aligns with business goals.

The DMM model helps organizations build a common terminology and shared understanding of how their data assets need to be managed. Its five successive capability levels provide a clear path for improvement in 25 process areas, reflecting all the fundamental disciplines of data management.

By providing a structured and standard framework of practices, the DMM can be leveraged by organizations to build their own roadmap to data management maturity. The DMM helps organizations become more proficient in their management of critical data assets, boosts support for tactical and strategic initiatives, and provides a consistent and comparable benchmark to gauge progress over time. It is a powerful tool to create a shared vision and terminology, clarify roles for all stakeholders, increase business engagement, and strengthen data governance.

The DMM may be used in whole, as an organization-wide evaluation baseline/benchmark, approached as a strategic initiative to establish or enhance the data management program. It may also be used in part, employing selected Process Areas in a defined or custom profile.

Many advanced solution projects – analytics, service-oriented architecture (SOA), data mining, Master Data Management (MDM), etc. – exceed their schedule and budget because sound processes are not in place when the project launches. For example, let’s say the organization’s top priority is to launch a master data management initiative for the Product subject area, critical to multiple lines of business (product development, inventory, orders, shipping, customer relations management, etc.) Let’s further assume that this organization does not have a robust operational data management program, therefore does not have many consistent practices and work products, and that this initiative is imperative to accomplish.

A set of the DMM’s Process Areas can be selected to target the most important fundamental data management practices for that need. In the above example, they would include Business Glossary, Governance Management, Metadata Management, Architectural Standards, Data Quality Strategy, Data Quality Assessment, Data Profiling, Data Integration, and a few others related to that specific goal. The list of Process Areas follows:

InfoQ: What made the CMMI Institute decide to develop this new model? Is there a need for a new model?

Melanie: Prior to its very first release, CMM addressed data management process areas as well as the software development life cycle. The Department of Defense determined that its emphasis was software engineering and development, therefore the content was not included. So it was part of the initial vision and has now come full circle.

Looking back 20 years, organizations in every industry were not ready to consider their data as a critical, valuable asset. With the advent of information engineering,  data warehousing, master data management, the ever-increasing volumes of data, etc., the industry has come to see that the data layer, and enlightened management thereof, was neglected for too long. Now, most organizations are more than ready to tackle what the DMM can help them to accomplish.

However, data management is broad and complex – so many processes, so many key work products. It is a challenge for any organization to fully address it, and has been a challenge for data management professionals to communicate to the organization the value of making a commitment to do a great job with its data assets, which are the lifeblood of its missions and functions. The DMM fulfills that need.

The DMM leverages the proven approaches, principles, and structure of CMMI, applying them to processes centered on effective management of data assets, which have been tested and proven over the past 30 years of data management evolution. It represents the state of the practice. The author team, which included more than 50 experts over the past four years, held itself to strict standards of model development.

We have done the hard work for the industry, producing a sound platform for any organization to learn, in a very short amount of time, exactly how they are doing, and what they need to do to improve. The DMM is highly useful as an accelerator for the program and an educational opportunity for every staff member who plays a role in building, maintaining, and governing data.

We think the need has been demonstrated by the many organizations and individuals who, when hearing that the DMM was going to be available, couldn’t wait to get their hands on it. It validates what the authors expected as long-term data-focused professionals - there was a hidden demand for a comprehensive measurement instrument that was practically useful and accessible to both lines of business and IT.

InfoQ: For whom is this new model intended? What are the kind of activities, roles and output that the model addresses?

Melanie: The model is intended for every organization that wants to more effectively manage its data assets. The companies that are already using the DMM are wide-ranging, including those from the IT, aerospace, financial and government sectors.   All industries, types, and sizes of organizations can benefit from the DMM.

An assessment against the DMM can be tailored to fit the organization’s needs. For instance, it can be applied to an entire organization, a single line of business, or a major project with multiple stakeholders.

It is also recommended to assess data management capabilities before embarking on a major architectural transformation initiative, such as implementing a Services Oriented Architecture (SOA). For example, there are many articles on the Internet that advocate an integrated data model as a precursor to beginning a SOA transformation. For instance, it is easy for the organization to plan and schedule a data service project, but when design begins, it may discover that there are multiple sources, no clear owners, no governance, etc. Needless to say, these deficits in capabilities will impact on-time quality delivery. When SOA is primarily technically driven and data management is not addressed, organizations may end up creating another data layer component (data services) on top of existing data layer components (interfaces), increasing rather than decreasing complexity. 

InfoQ: How can an organization deploy the Data Management Maturity model? Which benefits can they get from it?

Melanie: An organization employs the DMM by conducting an evaluation of its data management practices against the model.

Our five early adopter organizations – Microsoft, Fannie Mae, Federal Reserve System Statistics, Ontario Teachers’ Pension Plan, and Freddie Mac – involved key stakeholders from across their organizations to collaboratively evaluate their data management practices against the practice statements in the DMM. Individuals from lines of business, data management, and IT participated actively. They found that the group evaluation of the practices focused attention on accomplishments somewhere in the organization that others didn’t know about. They also realized that each stakeholder was passionate about their data, and gained the benefit of collective understanding about others’ priorities and challenges.

This educational effect is one of the most useful outcomes of the DMM in action. All of these organizations experienced a big boost in awareness and adoption for their data management programs, in addition to the comprehensive final assessment report, which precisely pinpoints gaps and provides certainty on exactly what needs to be addressed according to their objectives and priorities.

The DMM was intended to be a comprehensive a measurement instrument. Because it contains more than 300 practice statements and more than 300 example work products, it is sufficiently detailed that a typical organization ends up with 12-20 findings to address (i.e., practice statements which, if satisfied by activities and work products would support achieving the next level in the process area) as well as 8-12 recommendations synthesized from the DMM results, strengths, challenges, and priorities. Quick wins and tactical projects are discovered, as well as longer-term strategic initiatives.

It is perfectly possible for an organization to walk through the DMM with a few business, data management, and IT subject matter experts. However, we recommend broader participation, and take a “Big Tent” approach to our assessments with all relevant stakeholders in the same room. This maximizes understanding, collaboration, and agreements in a very short period of time. Work product reviews provide the evidence for the stakeholders’ consensus affirmations.

InfoQ: The Data Management Maturity model can also be used to bridge the gap between business and IT? Can you elaborate on that?

Melanie: The CIO, COO, CDO, and other business executives, frequently hear complaints around access to the right data at the right time for business purposes. Often data issues are blamed on IT; they deliver the systems which provide the data, so it must be their fault. And IT, in turn, expects the lines of business to take more responsibility for the data which they own and manage. When you’re driving a car, pedestrians worry you; when you’re on foot, drivers are viewed as dangerous. The push-pull-push back dynamic is a significant obstacle to progress.

The DMM was designed specifically to engender collaboration and evolve stakeholders’ perspectives to the synthesis of vision that is required for all components of the organization to fulfill their roles and share responsibility for the data assets. In action, the DMM’s practice statements and example work products, encountered and discussed collaboratively, result in unified conclusions. It interprets business considerations for IT and vice versa.

When explaining the DMM assessment to an executive, we are often asked “What if the participants get into an argument?” We’ve never had an argument. The nature of the material is conducive to encouraging people to make decisions together. Naturally, they tend to put on “the enterprise hat.” So in practice, the opposite happens – participants are energized and eager to share both challenges and achievements pertaining to the practice statements. The most frequent comment we receive is that it was very valuable to consider these capabilities together.

The DMM is not by any means an adventure novel; it is more like a textbook. If you read it page by page, you may appreciate it, but your heart won’t beat faster. It comes alive when it is used, and then participants can appreciate the knowledge of everyone else and recognize their commitment to their data. That puts business and IT on a new footing, and if the energy created is thoughtfully harnessed, the data management program will thrive.

InfoQ: Can the model also be used by organizations who have or are adopting an agile way of working? How?

Melanie: Yes it can.  Like the CMMI, the DMM is intended to be independent of any particular methodology.  The model states what should be done, the organization determines how to do it.

Agile methods can be used in support of the definition/implementation of the processes and there may be some practices that lend themselves to agile techniques, but the term “agile DMM” is misleading, because although data management is made up of processes and supporting work products, data per se is a “thing.” Also, DMM is an organization-wide-focused endeavor, and agile tends to be focused on small teams. The data assets are an essential infrastructure component, and just like the application layer, the data layer has typically grown ad hoc, project by project, over the years.

After a DMM assessment, the organization will know precisely what its strengths and gaps are. It then must choose its priorities – scope and disciplines - for managing data more effectively. For example, if improving data quality rises to the top of the priorities, an organization with that impetus may choose to emphasize those four process areas more than others initially. A sample priority for the data management program may be, e.g., “Implement a data quality improvement initiative this year for all the critical customer data stores.”

InfoQ: If an organization already uses the CMMI and/or People CMM, would it be interesting for them to also use the DMM. How can they combine these models?

Melanie: It would be interesting for them if they were trying to address enterprise data management issues.  There is almost no overlap between the DMM, People CMM, and CMMI -ACQ, -DEV, and -SVC and thus the models can function very well together where there is a need for the multiple disciplines. An organization may approach this by selecting specific process areas to evaluate. A partner who is familiar with a number of the models may suggest a customized set of process areas from among multiple models to meet an organization’s specific needs.

About the Interviewee

Melanie Mecca is an IT professional who has focused on enterprise data for more than 30 years. She has developed data strategies, architected and implemented data management programs and enterprise data architectures, and instantiated data governance and compliance for numerous Federal and commercial organizations. Since January 2011, Melanie has dedicated her efforts to the development of the Data Management Maturity (DMM)SM Model product suite.

Rate this Article

Adoption
Style

BT