Enterprise Resource Planning (ERP) implementation projects today are costing companies tens, even hundreds of millions of dollars, yet all too frequently they encounter major problems, delays and cost overruns. Worse, after months or years of spinning wheels and getting nowhere, more than a few eventually get cancelled outright.

Often, the initial thought is to blame the ERP package itself. But in most cases, the system, be it SAP or Oracle or whatever, is not the primary reason for the difficult implementation. Rather, the real problem can usually be attributed to poor process. How can I be sure? The evidence can be found in companies with superior implementation practices, which today are succeeding with the same ERP systems that don't seem to be producing satisfactory results for other companies.

What makes for a best-practices process? If there's one overriding principle, it's this: Treat the ERP implementation as a business-change project, rather than an "IT" project. That means: