What is the longest project you have ever worked on?
Mead Corporation’s implementation of the SAP ERP (Enterprise Resource Planning) system began in the year 2000 with a four-year timeline and a $125 million budget. Mead had looked at ERP systems a couple of times earlier in the 1990s and decided then that the company wasn’t culturally ready for a single, process-oriented system like SAP. But late in the 1990s, multiple divisions began requesting funds to implement their own ERPs, different ones, of course, and corporate had already purchased an ERP called PeopleSoft, so another look was taken and SAP was the ultimate decision.
The key focus of the project was change management, from executive management to division leadership to the employees that would transact in SAP every day. Four years is not a long time to implement a $4 billion company and executives tend to lose their enthusiasm towards the end. To jumpstart the project, a preconfigured SAP system was purchased from Monsanto, the chemical company, which was also in a continuous process manufacturing business. Eighty percent of what Monsanto configured for “Source and Support” processes, for example, accounting, finance, and purchasing was adopted by Mead and led to the first implementation only nine months after the project launched. That implementation was at the Coated Board division, selected because it produced paper, a critical proving ground to show that SAP would work in our core paper businesses, but also less risky than the larger and more complex white paper business. After implementing Source & Support, the project would follow up a few months later with the “Order Management” processes, for example, accepting orders, manufacturing and shipping the products, and collecting money. All this was supported by process training, local expert (“Super Users”) training, and finally just-in-time end-user training. We were within a year of finishing the project when the MeadWestvaco merger occurred, extending the timeline a couple of years, but at that time I took on the newly-created Chief Technology Officer role, and while still involved, I was no longer officially on the project.
My role during the project was the SAP Technology Director, having the BASIS, ABAP, and Data teams reporting to me, and I reported to the Vice-President in charge of the SAP project with a dotted-line reporting relationship to the CIO, my previous manager. The BASIS team was responsible for the installation and support of the core SAP software, the ABAP team wrote programs, interfaces, and reports for the team, and the Data team extracted, cleansed, and loaded data from legacy applications into SAP. I worked a lot with the I.T. organization and to meet the project’s aggressive timeline we had to use what we knew best, IBM mainframes, the DB2 database management system (DBMS), and Windows servers. The “usual” technology stack back then was an Oracle DBMS on large UNIX servers, but we had almost nobody on staff familiar with those, much less an expert. To support the large DB2 virtual memory requirement for SAP, Mead bought three of the first 64-bit mainframes that IBM rolled off their production line, and while that was a somewhat risky move, we encountered only a few manageable issues. We had to figure out how to manage an application with over 100,000 tables and indexes. We had to move to an integrated, once-a-month, four-hour data center outage, perhaps the most difficult change management effort. We had to create real-time, queued, and batch integration architectures and address locking and performance issues.
It was a very energy-filled, fast-paced, and continually challenging project which resulted in lots of great memories and lasting relationships.
Post a Comment