Configuration and Customisation

• Data Conversion Plan

1. Undertake work identified in the Data Conversion Plan.
2. Add any new areas that are found during the configuration and customisation.
3. Test the data conversion.

• Data tested and ready for a final conversion.

This activity should be undertaken as early as possible as it is typically one of the most volatile activities in terms of ability to predict the time and effort required. In addition, many of the corrections will require the skills of someone who is likely working on configuration and customisation, or user acceptance testing. It is best to get their time allocated early to the cleanup.

It is unlikely that all the data preparation activities will be identified in the Data Cleanup Plan. The identified activities should be undertaken, and when it is felt that the bulk of the work is done, a trial conversion undertaken. This will typically throw up another batch of errors caused by either errors you were not aware of, or requirements of the new system for particular data.

Decisions will need to be made as to whether some data is cleaned manually or a correction script written. There are basically two ways to look at this to make a decision as to which way to proceed.

1. What is the cost of writing a script to fix the data? What is the cost of manually cleaning it?
2. What resources are available to write a script? What resources are available to do a manual fix?

Whilst it may make financial sense to do it one way or another, the decision may come down to availability of resources.

When allocating tasks against resources, try to align the skills required for each correction with the skills of the person undertaking the work. For example, it is probably inappropriate to have a highly skilled person correcting telephone numbers.

Another point to enforce is that if a particular error is found by chance, it should be reported to the team leader. There may be similar errors in other records to be corrected. For example, if client name and addresses are being corrected, and someone finds where there were the postal address field had been incorrectly used to store the name of someone who was responsible for bill payments, it should not just be corrected for one client. It may indicate the field has been misused in other client records. A query should be run to look at all postal addresses for similar misuse.