[Technology Tapas] Migrating Data and Data Mapping – 4 Reasons It Can Be Challenging

Technology Tapas has been created to provide small, quickly digestible amounts of technical information geared toward the Business User who may run across the topics in a business setting and want to have a base foundation of understanding.

Tapas topics will take no more than 10 minutes to consume, and we will include Topic Discussion and Business Application, as well as an area for replies & requests, upcoming topics, and a link for those wishing to follow up with a subject matter expert.

Today’s Topic: Migrating Data and Data Mapping

If you are working on any sort of modernization project, moving data is likely smack in the middle of the effort.  In doing so, it is likely your business is looking at which data and how many years of data will bring the most value. As soon as you start asking “Which data is moving?”, you will have to know a little about data mapping. Data mapping is a very detailed view of the data at its source and where it will land (the target). As you estimate the work involved in your modernization effort, take a close look at this activity. It can be quite challenging and consume more time than you might think.  Here are 4 reasons: 

1. Legacy systems include valuable data. However, it may be that you have lost the internal knowledge around some of that data, especially if it wasn’t well documented. As people come and go, business logic may have been lost, making the mapping more difficult.  
2. Data mapping isn’t just about moving field A in the source to field B in the target.  Business rules matter, too.  Was the data transformed in any way?  If so, those business rules need to be considered.  Over time, a business may change the way data is represented – for example, a reorganization of business units – which may make mapping to the current organization structure difficult. 
3. Time references for transactional data may vary from system to system. For example, one system may contain only the final state of a customer order.  Another system may contain all the transactions that transpired for a customer order.  The mapping document would need to include the logic which identifies the current state of the order in the latter scenario. 
4. Moving transactional data vs data warehouse data affects data mapping as well.  For example, if the source system is transactional database and the target is a data warehouse, data may need to be aggregated in the target.  Data mapping documents would need to reflect the aggregation logic. 

At Perficient, we know that data mapping is an integral part of modernization and data governance initiatives, needing careful planning and estimation.  Click these links to find out more about modernization and data governance.