Branching Out Toward Data Management
For the rest of the July 2009 issue of CRM magazine, please click here.
Tell us about your organization.
Weyerhaeuser’s principally engaged in the growing and harvesting of lumber; the manufacturing, sales, and distribution of forest products; and real estate construction and development.
What were the drivers behind implementing a master data management (MDM) solution?
This all started with an enterprise data management strategy. The strategy outlines 10 steps of data quality. Step 1 is “Your data is unclean.” Step 10 is “Your data is accurate, homogeneous, and timely.” We were on about Step 2 of the data-quality scale and were looking for ways to improve. One of the ways was through MDM, to apply more-robust business rules against the data than could be done in a plain-vanilla implementation of SAP. We could have highly customized our SAP data-processing system, but that was expensive. Before ultimately choosing Siperian for its MDM hub, we looked at several other vendors and did a couple proof-of-concepts. We probably spent six months to eight months on the selection process. Part of that time was us trying to figure out what we wanted, too.
How’d the implementation process go?
We brought in contractors to do some of the programming work and we had one resource who worked offshore on some of the interface code, but we did all of the project management and the planning. We got the Siperian tool in-house and started our final version of requirements in September 2007; the build for the Siperian portion was done by the end of February 2008. That included additional things we needed to build, such as our user interface (UI) and integration into SAP. We tested it all the way through July, and went into production in August. One thing we did right was to have a very long testing window. While each piece could do its own testing, it wasn’t until you see the whole thing flowing through all the pieces that you see, “Hey, we have a gap here.” The more integration pieces, the longer your testing window needs to be.
How does the MDM hub work?
Our implementation’s probably a little different than most. A lot of people bring in data from the source system, harmonize it, and then spit it back out. Our approach is that MDM is a front gate to our SAP system. We make sure the data’s clean and then send updates to SAP.
In our UI, users start their request processes by searching for the intended customer. One of the problems we had before is that users would add a customer, not realizing it was already in SAP, and we’d get duplicate customer data. Now, using Siperian search and FuzzyLogic, if users find a customer record exists, they can update it. If not, the system takes them to a request UI, which starts the workflow process.
Throughout the workflow, we’re firing about 200 different business rules to make sure everyone’s informed. If someone enters data in the wrong sales area, the system automatically tells them it’s not valid.
What results have you seen and what are your goals with the MDM hub moving forward?
The data quality has gone way up and we don’t have nearly the number of problems that we used to. The data maintainers used to have to remember all the rules in their heads. We’ve also rolled out consistent processes for the end users. Before this, there was a variety of ways to get data to data maintainers—some had documents they faxed, some emailed, some called. Now there’s one way.
The real importance of MDM is for the downstream systems. If a salesperson puts in a wrong pricing code and they don’t get the right prices applied for their customer group, you have to go back and fix it. That’s not good for customer relationships. This system avoids the problem.
We already have a new release of our UI and workflow planned. We’re adding another group to the approval process and adding more email notifications.
FIVE FAST FACTS
How old is the implementation?
It went live August 2008.
Who was involved?
The business process and data owner, the chief information officer, and the director of enterprise information management.
What’s been the best decision during the past six months?
We’re replacing the front-end user interface, which will help cut out one of the different technologies. We want to consolidate for increased functionality and easier maintainability.
Address cleansing. Addresses need to be set to postal standards, so we implemented software that does that. After we went live, we realized that, since most of our customers are builders, most of our delivery sites aren’t postal addresses yet. We need to find a cleanse tool to work off geo-code.
We didn’t have enough end-user involvement soon enough. Because we were so focused on the implementation and migration to SAP itself, we couldn’t talk to end users about what they needed. They were brought in late in the game for testing. If we’d known when designing [the system], we probably could have met more of their needs with the first implementation.
Every month, CRM magazine covers the customer relationship management industry and beyond. To subscribe, please visit http://www.destinationcrm.com/subscribe/.
Informatica Snags Siperian for $130 Million
"This acquisition rounds out our MDM strategy beautifully," says Informatica's CMO.
Cegedim Dendrite Acquires SK&A Information Services to Expand Global Database
The purchase allows the life sciences CRM vendor to increase the breadth of its OneKey healthcare data.
Siperian's Last Dance: A Customer-Driven Module for Financial Services
With its Extended Customer View for financial services firms, the MDM vendor — acquired today by Informatica for $130 million — puts the emphasis on relationship-based insights.