This post was extensively updated and incorporated into my new book, The Data Conversion Cycle: A guide to migrating transactions and other records for system implementation teams. Now available on Amazon.com in both Kindle format for $4.49 and paperback for $6.99.
Many IT projects involve moving operations and users from one system, already in production, to a new system. Many of these are transaction-processing systems, such as financial management systems, customer relationship management systems, or human capital management systems. Typically, project teams will need to migrate data records from the system currently in production, which we’ll refer to as the legacy system, to the new system. This activity is typically referred to as data conversion. During the course of such an implementation project, there will typically be several conversions performed in support of prototyping and testing. Consequently, it is useful to think of data conversion as a cycle, repeated as needed.
The diagram below depicts a common model of the data conversion cycle. There are others, with more or less detail, but over the next few weeks, I plan to write about the various tasks in the model below. Note that my perspective on the subject is driven by my experience implementing enterprise resource systems, mostly human capital management, payroll, and benefits administration systems of the last twenty years or so. These days, I manage projects that involve migration to cloud-based, Software as a Service solutions, so I may make some generalizations that don’t fit all situations. So if you have any questions or objections, please note them in the comments. I’d like this to be a public collaboration.