I am asked quite frequently whether a company needs master data. Actually, I believe everyone has master data, it just depends on whether your company wants to manage their master data or not.
If I could start at the beginning (knowing what I know now) for a brand new company, I would design the data in way that master data is in ONLY one place, with applications making calls to that data. You know, all data for Customer, Product, etc., would be in one place. Data quality would be programmed in the insert and update processes based on ‘corporate’ business rules that would apply to this data across the board. But, we are faced with disparate systems with overlapping data, and data that may not of the best quality.
In fact, in the beginning of my data career, we were taught to NOT create more redundant data for the enterprise. Unfortunately, this lesson was lost as the various database technologies were taking root. Along the way, most enterprise groups were dismantled and absorbed in the company, and each new project created any data they needed to meet their requirements.
Hence, one of the reasons people created the first data warehouses – integration of disparate system data! Since most of us had “siloed” redundant data, we had to find a master data solution to feed, love and nurture new enterprise projects that required consistent good quality data.
In summary, your company may not need a master data solution, but it will need good quality data for consumption across the enterprise.