In a recent blog post called Understanding the sources of master data Prashanta Chandramohan writes:
“Often times, master data sources are legacy in nature, built and maintained over a long period of time, lack documentation and include procedures and terminologies which are no longer relevant in the current context.”
This saying resonates very well with my experiences.
Implementing Master Data Management (MDM) solutions doesn’t take place in a green field. Most of the hard work is not about how to build a perfect master data environment but is about how to work around what during the years has been done badly with master data for many good reasons.
Some of the maybe low practical but yet persistent challenges I have worked with are:
- Quite a few old systems hold data as names, addresses, product descriptions and so on only in upper case. You may want to convert that to a more beautiful mix of upper and lower case (according to the culture in question) on an ongoing basis. When handling master data entities describing things outside the English alphabet, we may even want to optimize the use of national characters in strings that before only allowed characters from the English alphabet.
- Fields are used for other things than the original purpose because there is no other way. While ongoing conversions including parsing may not be the best way around it often is the only way to go.
- Due to limited search capabilities in old systems you may write personal names starting with the surname (in cultures where that’s not common), twist company name elements around and so on. This may not look nice when mashing up with other sources and limit the use for other purposes, so also here conversions may the only way to go.
Please find some more on the fun in doing those things in the post The Cases for UPPER CASE in Data Management.