Data Cleansing
One of the most overlooked critical success factors of data migration, is early, targeted Data Cleansing. Whatever the project, once the initial scoping exercise is completed, which should clearly identify where any history is to be deemed out of scope and be signed off by the business/data owners, a comprehensive strategy for data cleansing should be developed.
The strategy will be developed in conjunction with the business/data owners and must contain specific KPIs that will be tracked on a periodic basis (suggested at least monthly). Once this has been signed off, it should be deployed immediately and the tracking metrics should be captured and base lined for the agreed KPIs.
An example of a KPI could be “open invoices older than two years” which should be actively closed; if no significant improvement is noted for the first few periods then escalations to the relevant responsible person should be made (the selection date would be rolling; this would need consideration before any escalations are invoked).
It should go without saying, that the older the data and more history that is migrated, the more complicated and prone to issue your data migration will be. If the data is not required for statutory or legal reasons or has no impact on core business operation, it should not be migrated. A read only archive of the non-migrated data will be made available as part of the overall project definition.
Cleansing data through the data migration processes should be avoided. Doing so puts the ownership of data quality on the project team, and also adds risk as well as creates a potential reconciliation and subsequent audit nightmare. Unless absolutely unavoidable, all data cleansing should be completed in the legacy systems.
Enisus can help you with defining the data project scope, writing the data cleansing approach, developing data cleansing KPI’s and the production of a Data Cleansing tracker (template available).