How to Maintain Master Data Management
There are a variety of ways that master data management (MDM) can be handled. No single approach is wrong but there are a few more common approaches that are addressed below.
Single copy approach
Here there is just one master copy that houses all of the master data. Whenever a change is to be made, it must be made in this copy. Any application that may be using the master data must be rewritten so that they use the new data instead of the old. This is a great way to maintain consistency in all applications, but in many cases it just is not practical. If you have to modify all of your applications to use new data, it can be very costly. If you do not own the rights to the software, you may not even be able to modify it at all, making the single-copy approach useless to your organization.
Multiple copy approach
Here the master data is still changed in a single copy, but the changes are then sent out to all of the applications that are storing local copies. This makes it easy for each system to update parts that have nothing to do with the master data, without making changes to the master data itself. For example, a payroll system might be able to change the salary for a specific employee, but it will not be able to add an employee. This will have to be done in the master copy. This approach reduces the need to adjust applications on a regular basis, but it gives only limited access to the data.
Continuous merge approach
Here the systems are allowed to make changes to their own copies of the master data. These changes are then sent back to the master copy, where they are then merged together. After a change is sent back to the master list, the master then sends new copies to all of the other systems so they may have updated information as well. The continuous merge approach requires very little change to the applications and virtually no application coding. There is no retraining required. All systems keep on doing their jobs and continue being updated as new information arises. Though it seems like a perfect solution, there are a couple of drawbacks. In the rare case that two systems update the same field at the same time, there is no way for the data management system to decide which should be kept. In this case, a person needs to get involved. There are ways around this problem but they involve extensive data governance rules and operating procedures. It can also be difficult to maintain consistent values. If an employee enters the weight of a product as pounds and it is converted to kilograms and then back again, the system will round the figure each time, potentially creating an incorrect number.
In most cases, all of these potential issues can be avoided with proper planning. The important part is deciding which way your organization should handle its master data management. There is no right or wrong answer, each company will have different issues and thus different solutions.