Business Intelligence (BI) is a rapidly growing method for understanding data to make informed decisions regarding company performance. In the past, few people had access to information to make decisions. This limited the ability for non-management employees to improve business processes and improve performance. Even managers who had access to the information had to wait for a business intelligence analyst to compile the data and provide them with a report, which created delays and made it difficult to make timely decisions that affected daily operations.
But that was the past. Today, few companies fail to see the value in BI to reduce costs, increase revenues, and mitigate risk or, for that matter, the point in keeping the information held within a tight group of managers.
Introduction to OLAP
As BI capabilities increased, new methods for analyzing data were created. One such method is Online Analytic Processing (OLAP). The method was first performed in a product called Express in 1970, but it was not introduced as an actual term until 1993 by Edgar Codd. OLAP gained popularity in 1998 when Microsoft released the first OLAP server called Microsoft Analysis Services.
OLAP is a very powerful method for processing data. In its simplest form it is a data structure designed to analyze multi-dimensional relationships to provide information. However, that definition does not explain the benefits of OLAP. There are a number of advantages achieved through implementing an OLAP structure.
First, OLAP stores data in a unique way. An OLAP “cube” is created with many relationships between the different data sources. A rudimentary example would be a three-sided cube storing revenue, time, and region. While it would be possible to create multiple tables to store this information, it would require as many as six separate tables (if you wanted to sort by each available category) to analyze the data in every way possible. And this is a simple cube with only three dimensions. OLAP cubes are not limited to only three dimensions, and each additional dimension would add a factorial of new tables to store the data using tables. Adding in just one more category would require 24 tables using traditional methods, and five categories would require an astounding 120 tables!
Simple 3 - Dimension OLAP Table
Compilation of 3 - Dimension Data in Tables
Second, OLAP queries can be quickly processed. This is a vital aspect, especially for operational business intelligence. The value of BI is to make decisions, often in a real-time environment, that impact an organization. If a query takes a lengthy amount of time, the value of the data is reduced because it cannot be acted upon soon enough. For example, if a call center is updated on call volume daily it is most likely too late to make a change to affect your service level agreement. Whereas getting call volume every few minutes allows managers to call employees back from coffee breaks or encourage employees to wrap up calls faster to handle someone on hold.
OLAP additionally provides a way to make data available to a broader audience. This helps to decentralize the decision making of the organization and allows for more ideas to be presented to improve the organization. While internally OLAP data is much more complex to set up and store, from a user point of view it is much easier to use.
An important point to consider about advances in BI, and OLAP in particular, is that the information is accessed by more people. Due to improvements in the user interface, queries can be performed without requiring IT support or extensive programming knowledge. This means that executives, managers, and employees can all make decisions based on the data. However, many of the people who now have access are not business intelligence experts or analysts. While the data may be available, unless it is presented in a way that makes sense to the user there is limited value, or in a worst case scenario, a faulty decision may be made because the data is improperly interpreted. This is especially evident as business intelligence is pushed out to the entire organizational hierarchy. Visualization is important, but more importantly, proper visualization is vital to communicate the information rather than simply to impress the viewer with graphics.