Asking the Right Questions
By Ari Kaplan
Data provides dry information on its own, but business analytics can pry the meaning loose.
Data is dead. It just sits there, taking up space in your database, on your storage system, useless. Sure, it has the latest inventory from the warehouse. But otherwise, it's of no help. It offers no insights, like the sphinx, just biding its time until someone asks the right question.
Business analytics can give you the right questions. It goes beyond ad hoc analysis and everyday queries and makes visualizing and analyzing business data easier; it can yield useful, sometimes invaluable, information about your business. An analogy might be the receipts you keep for the IRS: they're just slips of paper. But a skilled accountant, using standard procedures, can discover patterns of earning, spending, and investment that are tremendously valuable.
Business analytics commonly uses ordinary business tools to get answers. Online analytical processing, data warehousing, and decision-support systems all can be part of a business analytics suite. Modeling is also a part of business analytics, providing links between one entity and the others that affect it. These models are often dynamicchanging in timeto reflect the nature of business.
Paradoxically, the best type of analytical tool may not be one that simply answers your preconceived question. Instead, it may come up with the question for you, perhaps something you never thought to ask. This can lead to areas of discovery within your data that you may never have imagined.
Three common goals of business analytics are to understand what happened in the past, to see what's going on right now, and to predict the future. The right analysis can help decision-makers recognize what they need to do, or how they need to change, to reach their goals.
Oracle supports business analytics in several ways. First, it presents the data for analysis. This may involve data marts or data warehouses, cost-effective archives for historical data, and rapid loading of massive amounts of data. Second, it provides analytical tools. Finally, it offers ways to share information with the world, such as XML gateways.
With the acquisitions of PeopleSoft and Siebel, Oracle now offers a complete palette of tools in Oracle Business Intelligence Suite, which includes Oracle Business Intelligence Discoverer, Oracle Business Intelligence Spreadsheet Add-In, Oracle Warehouse Builder, Oracle Business Intelligence Beans, and Oracle Reports.
Specific tools address well-defined business areas. For example, look at the Siebel Business Analytics applications that are part of the Oracle Business Intelligence Suite. Oracle's PeopleSoft Financial Analytics applications provide the tools to develop, manage, and measure corporate strategies that can drive the most profitable growth. Each of these is like a question just waiting for the data to apply itself to.
None of this is meant to replace the normal queries and reports that businesses need. Rather, think of business analytics as adjuncts to the usual information generated, possibly providing new insights into existing data.
My work has used business analytics extensively in businessand in baseball. As a consultant to several Major League teams, I've helped decision-makers understand their situations and plan for the future. Business and baseball have remarkable similarities. For example, both sit on mountains of data and statistics that are just waiting to be used for analysis. Both also attempt to use past performance to predict future results.
Business analytics puts data in context. In baseball, you may have a player with 80 runs batted in. Is that good? Well, it may depend on how many runners were on base while he was batting, who bats after him, or other factors. In business, you may have a product that sold 80,000 units. Is that good? Again, the answer may depend on factors that business analytics can tease out for you. This process can give your enterprise some meaningful metrics.
When I worked for PricewaterhouseCoopers, my team analyzed products for Macy's and used advanced analytics to determine product placement. For example, some products sell better in some areas of the store than in others. Some products do well when near other products. Learning this kind of information buried in the usual business data can help any enterprise maximize its profits.
Finally, let business analytics answer a metaquestion about the data itself: is it useful? Some data can be very useful, because it's tied to trends or other important features. Other data may have no analytical use at all: it's incidental to the basic issues that decision-makers must consider. Thus, business analytics can tell us whether that data is uselessor just waiting for the right question.
Ari Kaplan ( firstname.lastname@example.org ) is president of the Independent Oracle Users Group (IOUG) and senior consultant at Datalink. He founded Expand Beyond, a leader in mobile IT software. He has been involved in Oracle technology since 1992.