What are the challenges in master data management

Master data management through the ages

Master data management between traditional requirements and new challenges

How do I convince the board of directors? How do I get the budget? How do I set up the master data management program? If like Dr. Wolfgang Martin has moderated an event such as the Master Data Management conference for so many years, so you have a lot to tell. For example, about what is changing and which topics keep recurring. Thanks for the interesting interview!

What has remained the same over time in terms of the topics?

Thank you, yes, I was the initiator of this forum around 9 years ago, and I have also moderated all master data management forums so far. Of course, I also had the opportunity to follow the development of master data management over this period.

A complex of questions that recurs and is always topical concerns the question of the business case for master data management. How do I convince the board of directors? How do I get the budget? How do I set up the master data management program? How do you organize ongoing operations? These are basic questions that have been asked and discussed in each of the events so far. In this context there is another question that has been asked again and again: How can you measure data quality and how can you evaluate it?

If you look at the topics over every 9 years, then organizational topics are clearly at the fore. Questions about technologies, architectures, tools and the selection of tools also keep coming up, but topics such as data quality and governance as well as centralization, globalization and standardization of master data are more important to the participants.

Not to be forgotten are the discussions that we had especially in the early years: Is master data management an IT task or a task for the specialist departments? This question has now been clarified in all companies: It is a joint task that is best cited in the sine of a primus inter pares by the specialist departments.

The motto of the upcoming forum is “Master data management between traditional requirements and new challenges.” What are these new challenges?

Two new challenges have already determined the discussion in the last two years: The question of cooperation in the cross-disciplinary and IT-overlapping teams using technologies and methods from social media and the question of whether master data management can be considered Can or should operate a cloud solution. Here you can now see the first companies moving in this direction. In 2013 Unilever presented its product master data management from the cloud, and in 2012 Hilti presented the possibilities of “Social MDM”.

This year there are two new trends in master data management. On the one hand, a technological topic: NoSQL databases in master data management. Here BMW will report on its experience in product master data management in vehicle construction. I see very clearly that such NoSQL databases can bring a technological boost to master data management, as they can not only manage the master data itself, but also the relationships between master data. This is ideal for processing parts lists, for example.

The second topic is again more organizational: It concerns the question of data share economy: opportunities and risks of joint master data maintenance. A reference to big data can also be established here, because big data analyzes need well-maintained master data. Or how else would you identify a visitor surfing your website as one of your customers?

Despite the stronger focus of the participants on organizational issues, the participants always discuss what makes the optimal software for MDM. In your opinion, what makes a good MDM platform?

I would like to limit my answers to the technologies of an MDM platform and not give a checklist for the functionality in detail.
A "good" MDM platform should enable central master data management, because master data is distributed across all processes and all applications. Therefore you need a central architecture for master data management. The best architecture here is a service-oriented architecture. By the way, the trend here is towards platforms with data virtualization. These are platforms that only manage logical master data and offer logical data services with the help of in-memory technologies. This eliminates the need for redundant data storage of central master data and the corresponding replication problems.

A “good” MDM platform naturally includes services for data quality management (profiling, cleansing, identity resolution) and for collaborative services (also in the sense of social media-inspired services) as well as a “good” search function. Data governance services are also an essential part of this. Further components are a rule and workflow engine that is separate from one another (or APIs for corresponding third-party products). Last but not least, administration and security are essential. This not only has to meet the legal requirements, but also has to be adaptable to company standards.

The data acquisition should be possible as a single acquisition and as a mass acquisition. An interface for data import and export is important, as is an API for product information management for the purpose of publishing catalogs (print, web, etc.). Nowadays, an MDM platform should also enable data entry, data display, reporting and monitoring for mobile devices and also be offered as a cloud solution.

And which requirements can possibly never be met?

Technologically anything is possible, it's just a question of price!

Thank you for the interview!

 

Author: Dr. Wolfgang Martin, http://www.wolfgang-martin-team.net

Contact: Tobias Knoben, Conference Manager EUROFORUM