The following is an extract from the mapping session of the EDI Training Course:
Mapping Best Practices
The following mapping best practices can be applied to most EDI applications.
Identifying and Understanding the Data
The mapping work should start long before opening the mapper feature of the EDI application.
The mapper should have a business analyst role of gathering data requirements. During the
process of gathering requirements, the input data format and the output data format must
be identified and understood by the mapper. The content of the data and the format of the
data have to make sense to the mapper. If the mapper feels that the input or the output
format of the data is unreasonable (e.g. too much unnecessary data) then the mapper should
make the case for this concern and not simply map away with unreasonable data formats.
This will help prevent troubleshooting and maintaining the map in the future and should
make the data conversion process more efficient.
Connect elements/segments at the same hierarchal level. For example,
connect header level segments on the input side to the header level
segments on the output side, detail to detail, and summary to summary.
Connect elements/segments with the same data type. For example, map the
numeric fields to numeric fields, date to date, string to string. Make
the data type in both the input and output section the same when possible.
Set the same maximum usage (loops) on the input and output side of the map.
Create all segments/elements in the map that will be in the file.
For example, if in the CSV file a certain column is not being used,
map it anyway. Some translators require all the fields be mapped
and this is a good practice for troubleshooting.
Use standards when possible (ASC X12, EDIFCAT, RosettaNet and etc)