Implementing a new application is a big task. It often involves modifying the current operating model and migrating the data from the current tools. For companies that have been around for years, one can imagine the amount of effort required. Data migration is a task that consumes arduous effort.
In order to do a good job, you need to understand:
- What the current data are used for
- Where the data are stored
- How important/useful the data are
- The relationship with other data
- The relevance of historical data
Knowledge on the above would help map data to the appropriate data fields in the new application and provide guidance on a logical approach to data transfer. This exercise is a crucial step. There needs to be clarity on what data are needed for the new application and the source for the transfer. Given that there is so much work to migrate data; does the business really need the old data?
Here are several questions one could ask to determine the value of data migration.
- What problems would arise if old data are not in the new application?
- Who would be affected?
- How often would you need to access the old data?
- Is there an acceptable approach to archive and access the old data?
- What is involved to look up the old data if they are not transferred?
- Is there a significant delay to pull old data from an alternate method?
- Would the delay be acceptable?
- Is it adequate to migrate a subset of the old data?
- What is the cost of the alternate method?
- What is the risk associated with not migration old data?
Evaluate the response to each of the questions and determine if it is necessary to do the data migration. No doubt that the ideal situation is to do the full migration. However, if the old data add little value or the alternate method to pull the old data on demand is adequate, it is worth pursuing and hence, avoiding the arduous exercise.
© Connie Siu 2013. All rights reserved.