Author: Andy Hayler, Co-founder and CEO, The Information Difference
Master data management (MDM) is by definition about dealing with data that is shared between different applications, whether this is customer, product, location, asset or other data. Information Difference surveys have found that large organizations typically have 15 different competing sources of master data; some have hundreds. If different applications use different versions of such data then problems arise, ranging from misplaced deliveries through to the business making poor investment decisions based on faulty data. For some years it has been possible to purchase MDM software that puts in place a central hub that will notionally resolve this problem. The hub either keeps track of the different versions and maps them together, or in some cases drives out unified “golden copy” master data back into transaction systems.
However, as early pioneers of such projects found, getting consistent master data is not just about technology. There was a good reason why all those separate applications have their own versions of data – different business users have different perspectives. A sales person cares about processing an order, but a credit controller worries about whether the order will actually be paid for. A marketer is concerned about whether a new campaign is successful, but a logistics person needs to know whether the products being shipped for that campaign with actually fit on a truck. These different views and needs, plus the complications of mergers and acquisitions, have led to the multiplicity of versions of master data. Human nature being what it is, once a particular line of business is in control of their data, they are reluctant to give it up. They may not trust “those guys” in another department to maintain the data properly.
Hence an MDM project is as much about internal politics and change management as it is about technology. If you are to standardise on a new global customer code then you actually need to convince, coerce or instruct numerous people around the organization to change their current way of doing things, and quite possibly give up control over “their” data. IT departments rarely have the authority to make a business department change what it is doing.
These issues caused many early MDM projects to fail, and have led to the rise of “data governance” initiatives. These activities, ideally led by the business or at least jointly led by business and IT, set out a management framework for the ownership and management of key data. There may be a steering group of senior people, with day to day activities driven by “data stewards” in the business, along with processes as to how shared data such as a global account code can be changed. There is usually a small group of full-time staff to drive the initiative.
An Information Difference survey of 257 global organisations found that 39% of organizations were implementing data governance alongside MDM projects. Two thirds of these were driven either by the business or jointly with IT. In such organizations, an unusually high 80% claimed to be measuring data quality (twice as much as is typical based on other surveys). An encouraging 67% had named individuals in place to resolve ownership disputes over data. Such initiatives do not run themselves: a median of six business and four IT full time equivalents were required to run the data governance programs. However, the extra cost and effort of putting in data governance programs seems to be paying off: 60% of those with data governance programs reckoned them successful and 19% felt they were very successful; just 6% reckoned their programs to be unsuccessful.
Certainly, my own practical experience of large-scale MDM programs backs up the notion that data governance is key to successful MDM projects. Few MDM projects fail these days to the technology, which is increasingly mature. The project horror stories that I have witnessed have mostly occurred due to in fighting between different parts of the business, unclear responsibilities or over-ambitious scope. These are all things that a good data governance program should be able to help with, and the data from this survey tells us that good data governance does indeed usually lead to successful master data management.
About the Author
Andy founded Kalido, which under his leadership was the fastest growing business intelligence vendor in the world in 2001. Andy was the only European named in Red Herring’s “Top 10 Innovators of 2002”. Kalido was a pioneer in modern data warehousing and master data management. He is now co-founder and CEO of The Information Difference, a boutique analyst and market research firm, advising corporations, venture capital firms and software companies. He is a regular keynote speaker at international conferences on master data management, data governance and data quality. Andy has a BSc in Mathematics from Nottingham University. He is also a respected restaurant critic and author (www.andyhayler.com). Andy has an award-winning blog www.andyonsoftware.com.