An ERP project has many stages. Change management is one of them. Data migration is another. For more on the overall ERP implementation methodology, read this article.
Definition and objectives
A data migration extracts data from a source system and loads it into a target system. In between, the data is cleansed and recoded. This means defining the scope of data to take over and formatting the source system’s data into something the target system understands.
In an ERP integration project, the source system can be another ERP, the company’s own IT tools, Excel files and so on.
The objectives of a data migration are simple:
- Thoroughly clean the data
- Give the target ERP relevant data
- Make key users’ work easier by building suitable extraction, recoding and loading programs
- Promote cohesion between business areas through shared data
Data to migrate
There is no fixed scope of data to migrate when integrating an ERP. It depends on the nature of the project and on the number of business modules to integrate. In most projects, however, the types of data listed below are part of it.
Master data
This is mandatory data that the ERP needs in order to work properly. It is usually migrated after the system is configured. It includes, among others:
- Suppliers (for the purchasing and accounting modules)
- Customers (for the sales and accounting modules)
- Items (for all modules)
- Replenishment rules (for the stock module)
- Supplier or customer price lists
- Bills of materials, workstations and product recipes (for the production module)
- Quality control ranges (for the quality module)
- Work equipment (for the maintenance module)
- Employees (for the human resources module)
Transactional data
This is the history of data in the old system. It is useful for business continuity and is migrated after the master data. It includes, among others:
- Stock levels
- Supplier and customer orders
- Supplier and customer invoices
- Production orders
- Quality control results
- Maintenance orders
- The accounting trial balance
- Employment contracts and payslips
- Employee leave balances and working time
Data-loading strategy
The guiding principles of the data migration stage are:
- A migration schedule must be set at the start of the project
- The resources needed for extraction, recoding and loading must be identified
- Only active, relevant data will be migrated
- Data will be “cleansed” before loading into the target ERP
- Only the historical information needed for mandatory reporting will be migrated
- Migration programs will be used for bulk migration. For manual takeover, the data will be created directly in the target ERP
A clear data migration strategy must be set before the ERP implementation project actually starts.
First, once the project team has listed the data to migrate, it draws up a detailed migration schedule, data type by data type (object by object).
Then, for each object, it states who is responsible for extraction, recoding and loading, and gives dates for each step. Because some data depends on other data, the schedule must be built carefully.
Next, for each object, it determines the volume of data to extract and the cleansing rules to apply. For example, for suppliers, you may decide to keep only those with commercial activity in the last 2 years, which means all the others are excluded from the load into the target system.
Finally, it determines how the data will be taken over. Loading into the ERP can be manual, or through a simple or more elaborate import program. The aim is to make users’ work easier.
Data migration in an ERP implementation is not a minor step. It is a key phase for the success of the project.
