The North American division of shipping behemoth A.P. Moller Maersk Group is reaping the rewards of a bold and large-scale enterprise data governance strategy, despite the ever-growing pile of IT industry research
In an era when most enterprise data governance efforts fail to make headway owing to a lack of executive support, funding and interoffice cooperation, Maersk’s multi-pronged approach combining people and technology is thriving, save for a few hiccups here and there, according to Jared Daum, the company’s director of master data management.
Daum said Maersk is planning to take the North American data governance strategy global. But how is a worldwide conglomerate with upwards of 115,000 data-generating employees achieving enterprise data governance when so many others have failed? According to Daum, it’s largely a matter of planning, collaboration, hard work and treating data governance like a never-ending process.
“It’s not a project that has a beginning and an end,” Daum said. “This is something that needs ongoing monitoring, and to do that efficiently, you need the right tools.”
SAP data integration plan leads to enterprise data governance strategy
Maersk Line – A.P. Moller’s North American Division – runs pretty much “everything.” That ranges from mainframes to Unix servers to blade servers running Microsoft SQL Server and Oracle Database, as well as hundreds of business applications. About four years ago, Daum was charged with overseeing the migration of Maersk Line’s data into an SAP R/3 Enterprise Resource Planning (ERP) implementation.
Daum’s job was to ensure that the data was in “good shape” for the migration, and that turned out to be a very tall order. The data being moved included information about products, prices, materials, customers and vendors -- and the records hadn’t been properly cleansed in more than 16 years. Daum needed to figure out what the SAP application required in terms of usable data as well as what Maersk-specific rules to set up to help ensure data quality in the future. Seven subsidiaries were being migrated to a single instance of SAP.
“We were sharing customers, and we were sharing vendors across business units, so how [could we] make sure that we had a common set of rules so that when one person changes something, you don’t have to play ping pong and change it back?" Daum said. "These were issues that we never faced before, and to make a long story short, I didn’t know.”
That’s when Daum decided to enlist some outside expertise. Based on a recommendation from Maersk Line’s CEO, Daum called in BackOffice Associates LLC, a South Harwich, Mass.-based software and consulting firm that focuses on data migration and enterprise data governance issues related to major ERP implementations. BackOffice began the process by conducting a comprehensive data audit and, according to Daum, the results were shocking.
The audit revealed that Maersk Line was plagued with duplicate files, incomplete data, invalid data, illegal characters, illegal lengths and other “stuff that just physically wouldn’t fit into SAP,” he said.
Maersk Line had about four months left before it was scheduled to go live with SAP R3, and something needed to be done quickly. The company decided that a combination of BackOffice software and a new approach to enterprise data governance would be the best way to help ensure that SAP R/3 went live as uneventfully as possible.
Combining active and passive enterprise data governance
To aid in the data migration effort, Daum and his team decided to purchase BackOffice’s Data Migration Toolset, which, among other things, helped the company ensure that information being moved was SAP-ready. But a successful data migration was only half the battle.
Daum also needed a “passive data governance” tool to enforce data governance policies long after SAP went live. Maersk accomplished this by purchasing BackOffice’s Data Dialysis software, and Daum says it’s currently doing a good job of finding the problems that humans tend to miss.
He has been very pleased with BackOffice Associates, but he has occasionally had trouble getting Maersk’s higher-ups to appreciate BackOffice’s offerings, because they run on Microsoft SQL Server.
“In the Maersk organization, Oracle and IBM DB2 are what we would call the big irons that we use, so when [the higher-ups] look at SQL Server, they think that it’s more of a kiddy toy, even though I think that is a false perception,” Daum said. “I have a little bit of difficulty convincing them of the value of some of the things we do because they just look at it and say, ‘It’s an SQL Server solution, so I don’t think so.’”
Maersk also decided to form an active enterprise data governance committee consisting of workers from various business units. The group’s role is to identify, implement and enforce company-wide data governance policies. Each member also spends time monitoring the quality of data entering the SAP application.
As a result of its ongoing data governance process, Maersk Line has reduced the number of erroneous vendor files to 1%, Daum estimates.
“We’re continuously working to try to pound those remaining ones out,” he said.
Beating the enterprise data governance odds
Maersk Line’s success with enterprise data governance is far from the norm.
According to a recent SearchDataManagement.com survey of technology professionals, enterprise data governance faces major obstacles, such as a lack of support from the executive ranks, corporate politics, and highly siloed business units. As a result, most companies have no formal enterprise data governance strategy in place. And data governance initiatives that do exist rarely extend beyond the IT department.
The failure of many senior-level executives to see the value of enterprise data governance can ultimately lead to problems, such as expensive data migration initiatives in the future, according to David Loshin, president and principal data management consultant of Silver Spring, Md.-based Knowledge Integrity Inc.
“I think the business leaders need to understand that linking businesses’ value to the use of information is a necessity today,” Loshin said. “Believing that your business runs separate from the data that it runs with is a risk that business leaders should be more wary of taking.”