Originally appeared in MSI Magazine
February 1, 1998

PDQ ERP

Learning from experience should be elementary. While necessity may be the mother of invention, experience makes those inventions successful. So, while the need for warmth drove early man to rub a couple stones over a woodpile, experience revealed that such techniques were terribly efficient or reliable. Further experience observing the workings of nature gave birth to science and technology—and in this case, more reliable furnaces that burned more efficient fossil fuel.

Manufacturing systems evolved in similar fashion. After hundreds of years of trying, manufacturers eventually codified materials management processes to avoid shut-downs of production lines because of inventory outages. Shortly thereafter, MRP software emerged to help manufacturers implement those processes more consistently.

Although things have changed a bit since the days when manufacturers worried about running low on inventories, the need to manage manufacturing processes hasn’t. In fact, while MRP has metamorphosed into ERP, its reputation for traumatic implementation has remained another constant. The horror stories of projects that went well over their time and materials budgets are practically folklore. Has the school of hard knocks helped anybody here?

According to SAP, Oracle, Cambridge Technology Partners, HP Professional Services and others, implementing on time, on budget is a snap—as long as you follow their script. Leery of falling into black holes, manufacturers are taking these ideas seriously. We examined a half dozen manufacturers of all sizes to check the viability of “plain vanilla” ERP.

Nobody admitted to us that they changed anything—almost. And nobody admitted that they strayed from time schedules or budgets. But they were eager to demonstrate their vigilance in maintaining project control.

For instance, a small, fast growing 125-person manufacturer of computer peripherals implemented Oracle Manufacturing with a seven-person project team in only 70 days—20 days ahead of schedule, and under its $900,000 project budget. The operating philosophy was avoiding all changes except field lengths, restricting project team to director-level staff, and imposing a process where silence on any issue was deemed a “yes” vote. Evidently, the company was leery of putting users onboard to avoid the temptation of kluging Oracle to look like the old application being junked.

We also examined a $450 million midwest agricultural products manufacturer which kept its R/3 process in check by stationing it three internal project teams at the regional facility of HP Professional Services, the systems integrator, and provider of project and functionality templates. In other words, keep the project folks away from the temptations of the home office, where they might otherwise get distracted by wish lists.

The idea is that, by using templates developed by software providers or implementation consultants based on prior experience, project lengths and costs can stay under control, while project leaders can keep their jobs. We found that with midsized manufacturers, that’s certainly possible, but it takes draconian measures to accomplish it.

Probably the fairest analogy is the compulsive dieter, where the pantry must be stripped of goodies in order for the diet to succeed. That was the strategy employed by a Fortune 250 chemical and pharmaceuticals giant which has plowed several hundred people into an R/3 project, which will ultimately last five years and cost several hundred million dollars. Control is maintained by a policy of turning down change requests, coupled with a replicated central repository of definitions and transactions to enforce consistency.

Is there any such thing as plain vanilla ERP? Most answers boiled down to, “Sort of.” “Configurations” where functions are selected from the vendor's palette, or where fields are slightly modified, were OK. But adding new functions were not.

We found few instances where manufacturers were able to hold the line 100%. The agricultural products manufacturer greatly revamped R/3’s order processing because it works on annual, not monthly order cycles. Furthermore, its product return policies were far more liberal than most manufacturing sectors. Another organization revamped the general ledger of PeopleSoft with a dummy customer listing because the vendor package could not easily consolidate revenue flows between the organization’s numerous subsidiaries.

So, the next time a you examine a bid for a fixed-price, fixed schedule ERP project, map your boundaries, then check on the consultant or vendor's definition of “plain vanilla.”