User:Cptkaveman/sandbox

Source independent format As with many data warehouse implementations, there's no shortage of raw data. The data in most cases needs to be linked with other data sets within the data warehouse. In some instances, data can be at varying levels of detail. When designing a custom database, it is important to understand this fact and preparing for it ahead of time, will save you a lot of time in the end.

Understanding the data What types of data will you be storing? What is the most granular (atomic) piece of data? Are there intermediate summaries of this data? Where will the data be sourced from? If multiple sources, do they use the same identifiers throughout? Are they at the same level?

Let's say you are storing data related to assets under management. In this case, you may have data at a product, distributor, client or portfolio level. For example,

Product level - A company may track the combined assets of a mutual fund. Distributor level - Assets of a product sold through various distributors (banks, financial institutions, etc) Client level - Assets held by clients in a specific product Portfolio level - Assets held within clients portfolio(s) or account(s) of a specific product

Separate topic Do the details from one tie back to the other?