Quality built in, not bolted on
In IDN, profiling, cleansing rules and de-duplication happen before conversion, in the same flow as mapping and loading. Bad data is dealt with before it reaches S/4HANA, and the same rules run again automatically in every cycle.
| The project way | Cleansing is a separate effort whose problems surface late. |
| The IDN way | Profile, cleanse and de-duplicate before conversion, every cycle. |
The problem
Cleansing is often treated as a side project. Duplicates are merged in one place and break related records somewhere else, and the same clean-up is repeated by hand in every mock cycle.
How IDN does it
- Profile: completeness, valid values and duplicates per object.
- Cleansing rules: written once, approved, and applied automatically in every cycle.
- De-duplication: duplicates merged with a crosswalk that dependent objects use.
- Approval: source data is approved before it moves on to conversion.
What it means for the program
Cleaner data reaches S/4HANA, and cleansing effort goes down with each cycle instead of repeating.
Questions
How do you assess data quality before an S/4HANA migration?
Profile each source object for completeness, valid values, duplicates and broken references before mapping starts. IDN shows the findings per object, so cleansing work is planned rather than discovered in test cycles.
How do you remove duplicate master data before migrating to S/4HANA?
Find likely duplicates, agree the surviving record with the business, and keep a crosswalk from old to surviving numbers. Records that depend on the merged ones, such as bills of material, must use that crosswalk or they will fail later.
Should data cleansing happen before or after conversion?
Before. Cleansing rules applied to the source, and approved by the business, mean the same clean-up does not have to be repeated in every cycle.