Maximizing accuracy in large data conversions

Data conversion looks easy to a lot of people. It's not-particularly if it's a large project. There are just too many ways things can go wrong. You can lose data or mess it up. And you can stall the conversion if you don't have enough computing resources. Project size really makes a difference. Some type of mapping tool can improve your chances for success because the data in the specification is the same data that drives the program that loads the target tables. Automation eliminated the mistakes that cause inconsistency. The tools and processes we developed on a large project worked well on a smaller project, but the reverse certainly would not have been true. A systematic process is required with some way to validate each step.