LAS VEGAS -- A traditional master data management (MDM) paradigm is that "process, not data, is key" to success.
But in recently completing such a project and going live with SAP Master Data Management, Kraft Foods Inc. found that needed to be changed.
"Process and data are critical," said Sue Barber, business systems manager for the company.
A successful MDM project requires attention to both, particularly to data governance, as well as people with expertise in both ERP and MDM. The Northfield, Ill., food giant selected SAP MDM in July 2006 and went live with it in September 2007.
At SAP TechEd last week, Barber shared some of Kraft's keys to a successful SAP MDM project and go-live, which happened within a year. The company created four repositories -- products, materials, customers and vendors.
Kraft encompasses multiple companies with its product lines and each had consolidated its data sources without harmonizing them, so that there was no underlying lowest common denominator. Additionally, different processes and tools were used to manage master data. Aggregation, a process in which data is gathered and expressed in summary form for statistical analysis, compounds data discrepancies at lower levels, Barber said.
"We also wanted a single source of the truth," said Barber.
Having a "single source of the truth" would help Kraft make better decisions, improve collaboration with suppliers and trading partners and more easily comply with reporting requirements, she said. The company wanted the project to improve its data quality.
Kraft also wanted a unified data foundation for its SAP ERP rollout.
SAP ERP customers typically opt for SAP MDM, according to a recent Gartner Magic Quadrant report on MDM for customer data. But SAP would not provide data on how many NetWeaver MDM customers it has.
The MDM project team at Kraft wanted to group all the corporate data maintained locally in dozens of locations around the world according to product data, supplier data and customer data.
They identified the data, largely with the help of the governance teams, selected the data management tools and determined the business process for creating, maintaining and using the data.
They reduced the number of data sources, increased the commonality of data and created consistent attribute definitions, Barber said. To manage bad data, Barber said they worked on standardization of the data so that they had a harmonized record.
And, most importantly, the new system supported a three-tiered governance model -- data governance at the global, regional and local levels. It was one of the most critical components in paring down and getting the right data.
"We tried to be really smart about what data we needed, and governance played a role in that," she said.
She said one of the most important elements is ensuring clarity on data object ownership within ECC. Complex business processes must be well-defined, she said.
"Don't underestimate the business change the system changes will create," she said.
When integrating ECC, make sure there are people on the team with expertise in both disciplines, Barber said. She said to use technical consultants who have created guided procedures and workflow. The team should be made up of people with experience in SAP MDM configuration and modeling, ECC 6.0 and process integration, as well as business analysts and project managers.
"When integrating ECC, expertise in both MDM and ECC is critical," in such areas as security enhancement, for example, she said.
Kraft is using NetWeaver Process Integration as the integration layer for moving foundational data into the SAP ERP application. It's also implementing universal data feeds out of MDM as part of its service-oriented architecture, which allows SAP and non-SAP systems to get the data without building multiple interfaces.
So far, end-user feedback has been positive, Barber said, namely because it's the same look and feel across the different data domains. They're also working toward an SAP portal user interface, Barber said.
Thinking outside of the box on design, and showing a prototype to the end users, really accelerated the buy-in, she said. And talking to other SAP MDM customers helped them figure out that they wanted SAP.
"It's a good way to get the business side to buy into it," she said.