Foundation: The MDW Model

Overview: The MDW Model

A Model for Today's Business Needs

As mentioned previously, the Dimensional model has been tried and proven over the last few decades to be one of, if not THE most effective database model for Analytics. However, when this model was designed the industry looked very different. Data storage was very costly and the tools and people analyzing this model were very different. In today's world storage is cheap and the average business user has access to powerful self-service analytics tools like Power BI, Qlik, and Tableau. In addition, the automation behind our platform eliminated the time and money constraints confronting traditional methods. Finally, the full architecture of TimeXtender, means the need for simplified navigation is handled in the Semantic Layer. For these reasons, we have designed the Modern Data Warehouse (MDW) to address these key differences. We still wanted to leverage the key benefits from the Dimensional model such as the resistance to change and optimized query performance. However, we knew that today's self service user demanded more than what a traditional star schema could provide, and automation made the change easy.

More Fields

The MDW starts with the traditional star schema, but adds additional tables and fields that were not present in the old model. Traditionally architects would remove fields from the final schema, bringing only what was necessary for their reports. This addressed the constraint on expensive storage capacity and traditional reporting methods of the last few decades. However, we've found it's actually more beneficial to not remove any fields when bringing tables to the presentation layer. Automation and inexpensive storage means this is cheap and fast. This also means that self-service users now have more data to explore and power their decision making.

More Tables

The MDW also includes more tables than the traditional star schema. While they may not fit into your typical model, transaction header tables and reference tables can provide many insights and answers to self-service users. The majority of business questions can actually be answered faster by querying the transaction header tables directly. Also, reference tables can provide valuable insight into flags or business definitions.

More Answers

In closing, the Modern Data Warehouse provides more answers, is more agile, and more resilient to change to serve the needs today's self-service users.