As you've probably seen, this month's theme is "BI and OLAP." To those acquainted with the specialized technology related to business analysis and databases, OLAP is a useful organizational structure of data. But to those who wouldn't know OLAP from a naughty Irish dance, this article will give you the raw basics of a potentially complicated subject.
First off, OLAP stands for "online analytical processing." Some consider it a set of protocols, while others may call it a technique or approach, but basically it's a flexible way for you to make complicated analyses of multidimensional data - and it provides this information with rapid execution. Okay, let's back up.
Consider your business's sales revenue as a measure (i.e., it's a KPI element that represents a numerical value). When you categorize that measure with something else (by product or by date, etc.) you create a dimension. Multidimensional data is at the heart of an OLAP system, where you group your data by more than one dimension (e.g., sales by date and by geography).
Furthermore, you are probably familiar with relational databases - all organizations employ these applications in one form or another to store their business data, be it financial, customer, personnel, logistical - anything and everything! The system usually organizes this data in two-dimensional tables, with rows and columns. Think of a familiar spreadsheet.
"OLAP tools store data in multidimensional databases or cubes, which are like spreadsheets on steroids - supporting multiple dimensions instead of just two," reports industry expert Wayne Eckerson.*
When data is stored in multidimensional cubes, it opens many possibilities for selectively looking at the information from a number of different points of view. Now you can more aggressively filter/sort data and drill up/down through it; you can explore, navigate and refine data until a desired snapshot is achieved. Practical examples include looking at sales numbers and asking: Why was one month better than the previous month? Which product was the top seller per department? And are staff levels in each department making a difference?
Sales information is just one broad example. OLAP tools can be immensely helpful for data mining projects, business process management (BPM) applications, budgeting, forecasting, planning and more. In fact, OLAP is a critical part of all modern BI technologies. When you give yourself an infinite set of data views - with varying levels of granularity - you can reveal information that would otherwise be difficult to attain. This maximizes your data-analysis potential.
"OLAP tools make it easy to uncover the root causes of problems, identify trends and compare performance across groups," Eckerson states.
Not only does OLAP provide the power to make complicated analyses, it does so in a relatively small amount of screen space, making it particularly useful for digital dashboard solutions. Furthermore, applications employing OLAP offer end users the ability to quickly achieve the exact data view they require. This makes managers more self-sufficient - and by removing their dependence on IT personnel, they are free to go about their work without asking software developers for help.
These are some of OLAP's main benefits. Of course, it should be noted there are several flavors of OLAP. Variations on the name include MOLAP (multidimensional-array OLAP, the "usual" form), ROLAP (relational-database OLAP), HOLAP (hybrid OLAP, for both types of database structures), WOLAP (web-based OLAP), DOLAP (desktop OLAP) and RTOLAP (real-time OLAP).
In sum, as organizations grow, so does the size and complexity of their data. This makes the task of manipulating and visualizing the data increasingly difficult and time consuming. To mitigate the risk of data becoming unmanageable, comprehensive data analysis and visualization solutions are required. By utilizing OLAP technologies, BI professionals maximize their data analysis potential via a significantly enhanced set of data views. This gives them the ability to transform data into valuable, useful information in a highly flexible and customizable manner.
* Source: TDWI Best Practices Report "Beyond Reporting: Delivering Insights With Next-Generation Analytics" (Q3 2009), p.18.
About the Author
Rob Hunter works as a software copywriter by day and as a Dashboard Insight editor by night (when he’s not playing his upright bass).
(Copyright 2010 - Dashboard Insight - All rights reserved.)