It’s a fact. In-Database Analytics dramatically decreases time, effort and resources required to generate analytics-based reporting. So when do you know it’s time to implement In-Database?
You’re ready for In-Database Analytics if you experience the following:
- Increasing data volumes, data velocity, data structure complexity. Although these alone don’t mean your organization should implement In-Database Analytics, they often signal a need. Increases in data volumes and structure complexities can slow down analytics.
- A need to reduce turn-around time for analytics. You’re not receiving your analytics-based reports quickly enough to maintain competitive advantage. Migrating analytics away from desktops and servers and into the Enterprise Data Warehouse (EDW) reduces analytics turn-around time.
- A need to audit analytics. Have you ever wondered how a piece of analytics was conducted? And when you asked, the explanation just made your eyes glaze over? With In-Database Analytics, all data prep work and most analytic steps are stored inside the EDW or within instructions between the analytics server and EDW.
- Large desk-top data marts or repeated data dumps to desktops. This is probably the most telling indicator for the need to implement In-Database Analytics: when analysts dump large data volumes to their analytic servers – again and again. Analysts use these dumps to perform data prep and analysis. These data-dump activities waste valuable time, put strains on the EDW and networks and often, in and of themselves, lead to all sorts of analytic dangers. By moving the data prep stage away from desktops and into the EDW, hours of work are reduced to seconds.
- FTEs devoted to building and maintaining analytical data marts. If employees are devoted to building and maintaining data marts used by analysts, it may be time to re-think this work effort. Instead of building and maintaining analytical data marts outside of the EDW, it is easier and more efficient to maintain an analytical semantic layer inside the EDW. Maintenance of semantic layers consumes less time and fewer resources than analytic data marts.
There may be many more reasons for implementing In-Database Analytics, and you may not necessarily agree with some of these, but what is undeniable is that implementing In-Database strategies reduces the work effort and increases the stability of analytics.