How to Get Quick Results From Your Master Data Management (MDM) Program

Jake Freivald's picture
 By | februari 22, 2017
in Omni-Gen, business process redesign, change management, City of Brampton, Dan Power, data governance, data integration, data mastering, data monetization, Data Quality, data quality artifacts, golden record, Hub Designs, Information Builders’, IT, jake freivald, legacy investments, Master Data Management (MDM), metadata, model-driven, omni-gen, organization design, pros, reengineering, remediation, risk assessment, St. Luke’s University Health Network, State of Louisiana, third-party data customer segmentation analysis, Wayne Eckerson, Master Data Management
februari 22, 2017

Last week I had the pleasure of co-hosting a webcast with Dan Power, the founder and president of Hub Designs and an expert on master data management (MDM) implementations. Part of our 2017 Innovation Sessions, we discussed five things you can do to fast-track MDM and data governance projects. I’d like to share a few observations from the webcast and encourage you to tune in for a playback of the complete session.

While many organizations put MDM projects in the hands of IT pros, Dan outlined an inclusive strategy in which the business community drives the project and IT plays a supporting role. He went on to describe the five essential elements of MDM success:

  • People (including organization design and change management)
  • Process (business process redesign or reengineering)
  • Technology (MDM hub, data integration, and data quality)
  • Information (both internal and third-party content)
  • Data governance (to get off on the right track and guide the initiative in the future)

If your MDM program doesn’t include all five of these elements, you’re increasing your risk of failure substantially. It’s a five-legged stool, and removing one of the legs means your project is more likely to collapse.

People

While IT pros tend to obsess about the technology that automates MDM projects, Dan reminds us about the importance of “the people stuff.” Whenever an activity crosses multiple boundaries within a large organization, there are issues to resolve – cultural, political, and most importantly, functional issues about what data should be mastered and how it will be used. He recommends identifying well-connected business leaders who can drive the project, keep senior management engaged, and allow the business to own the initiative while IT acts as a supporter and facilitator.

MDM strategy should align with corporate strategy and be explicitly tied to corporate objectives. What does that mean in practice? Spell out exactly how master data will support your company’s goals in the years ahead. Show people what’s in it for them. Explain how their business interests align with the goals of the MDM program.

Process

The heart of the webcast describes how to create a repository of high-quality master data as a foundation to improve business processes. This doesn’t happen automatically. You need to redesign your processes. Ideally, improving your master data should precipitate process changes, and process changes will yield better master data.

The MDM hub is where you bring together all the data from your source systems and define “golden records” using business rules. Implementing this hub requires careful consideration of the process improvements, and the associated master data curated within the hub.

Dan discussed the importance of having a model-driven approach in which business analysts build models that drive the implementation process. You can generate an entire MDM application from the models and metadata, he explained, which is captured by business analysts and end users. It’s an iterative approach that begins with a prototype and gets progressively better as the model approaches production.

Technology

The right technology platform can guide and synchronize these efforts – from initial idea assessment to final production environment – while structuring data quality, remediation, and governance. Technology helps you get data from your source systems into the MDM hub, as well as to transform it, model it, and generate a working MDM environment. Ideally, your technology platform should help you collapse multiple occurrences of entities, such as “customer” or “product,” into golden records, so you have something of value to publish back out to the rest of the enterprise. More about this in a moment.

Information

Data doesn’t just come from your internal source systems, but often from external sources as well. Don’t forget to pull in relevant third-party data attributes such as industry codes, revenue, age, number of employees, corporate hierarchies, financial risk, and other variables. Otherwise, your customer segmentation analysis will be flawed, or your risk assessment will be incomplete, or your sales roll-up by corporate family will be wrong. Look ahead to reporting and analysis and consider all the attributes you’ll need to support your BI initiatives.

Data Governance

A strong data governance function is essential to keep your analytics project on track. Dan says you should design and build your data governance foundation at the outset. If you take time up-front to design data governance into your initiative, your chances of success go up, because you’ll be drawing on the expertise of a larger set of people in your company.

Automating the Effort

My goal during the webcast was to describe how you can use Information Builders’ Omni-Gen platform to guide MDM initiatives. Omni-Gen coordinates many overlapping activities, from identifying the right data to describing what golden records look like to defining mapping rules, and finally, helping people envision how analytics can be applied to meet the original business goals. It’s a holistic approach that keeps the business firmly engaged throughout the process.

Omni-Gen helps you define and generate an MDM hub that contains all of the pertinent data. And it generates the integration, mastering, and data quality artifacts – including mapping rules to pull in your data and enforce the data quality rules that you have defined – all in a fraction of the time that such projects typically require.

Omni-Gen aligns business people with IT people, and it generates applications that combine data integration, data quality, master data management, and data governance to fast-track MDM projects.

For example, St. Luke’s University Health Network used Omni-Gen to help us develop Omni-HealthData. This anchored an agile MDM project, in which business professionals were able to simultaneously focus on creating new reports while mastering the data. This unified effort enabled St. Luke’s to launch an initial set of dashboards in just four months.

Omni-Gen allowed the State of Louisiana to preserve its legacy investments and more fully use its portfolio of data and information assets. The state has created a repository of clean, consistent data that can be shared by multiple agencies.

City of Brampton built a centralized master database for all properties in the city, using the same technologies that now comprise the Omni-Gen solution. These technologies monitor the influx of data and deliver it into a master repository, while correcting data quality issues.

Take a look at the webcast, now available on demand, to learn more about the unique technology and approach that these customers employed to complete their MDM projects and ensure successful analytic outcomes. And don’t miss our next Innovation webcast, where Wayne Eckerson will discuss how to monetize data through analytics.