Enterprise business performance management
Business intelligence + data warehouse = optimal business performance

by Rich Clayton, vice president of product marketing, Hyperion

In business, some questions are universal: How can we increase sales? How can we maximize margins? What areas of the business are growing the fastest? What performance metrics should we use to create incentives for our employees?

Other questions are industry-specific. A bank needs to know which of its branches are most productive. A retailer needs to determine the best way to reduce inventory costs. A manufacturer has to align production with demand.

But regardless of whether the questions are universal or industry-specific, decision-makers need answers. They need those answers quickly. They need them to be accurate and based on up-to-date information. They also need the answers to their questions to be easily accessible.

In fact, businesses that can quickly and easily deliver accurate, timely answers to decision-makers will consistently out-perform those that can't. They'll allocate their financials and employees more wisely. They'll respond more quickly to emerging market opportunities and abandon unprofitable activities earlier. Every manager at every level will be more empowered to make better decisions more often.

This is the value proposition that makes overall business performance management and business intelligence (BI) an essential requirement in today's fast-paced, hyper-competitive business environment. And, for companies with robust data warehouse environments, that value proposition is even more compelling.

Leveraging the warehouse

Data warehouses play a central role in any enterprise BI architecture. They ensure that data from all key business applications is available in one place for effective reporting and analytics. They can also ensure that this data is updated on a real-time and/or near real-time basis as appropriate.

However, to fully leverage such a data warehouse across the enterprise, IT organizations have to fulfill some fairly challenging BI requirements. These include:

Scalability of performance
Data warehouses can get very large in terms of both raw size and number of dimensions. In addition, as business users experience the benefits of BI, more of them execute more analytical queries more often. The complexity of those queries also tends to increase. All of this puts significant strain on the BI platform. It is therefore essential to implement BI technology that can scale to maintain performance even as the demand for BI grows. Otherwise, success can rapidly turn into failure as performance deteriorates and users can't get the answers they need.

Diversity of demand
Different users across the enterprise tend to ask different types of questions. A finance manager, for example, may need to apply complex cost allocation algorithms to thousands of products to understand their relative profitability. A marketing manager, on the other hand, may need to analyze millions of customer histories to discover if specific demographic attributes are linked to specific buying behaviors. These two analytical exercises are actually quite different from each other from a technical perspective. An enterprise BI architecture must be capable of addressing these diverse types of queries in order to meet the needs of users across the organization.

Interfaces and visualization
To take full advantage of the insight that BI can provide, users need a variety of analytic interfaces. To ask the right questions, they need a graphical interface that makes it easy for them to construct appropriate analytical queries. To fully understand the answers, they need presentation formats that visually highlight the resulting data sets. In some cases, it's best to create desktop "dashboards" that consistently provide executives and other decision-makers with real-time visibility into key performance indicators.

Other factors may also influence the choice of a BI platform. Cost efficiency, for example, can help lower the overall cost of BI ownership by delivering requisite scalability with a relatively low investment in hardware infrastructure. The integration of a BI platform with an organization's data warehouse platform of choice and/or or other potential data sources is also a key consideration.

Regardless of selection criteria, however, the importance of a scalable, flexible and easy-to-use BI platform to business success should not be underestimated. When decision-makers of all kinds can make optimum use of the data warehouse via BI, the entire business wins. IT in particular also wins because it realizes significant additional value from its sizable investments in the creation and ongoing management of the data warehouse.

Making BI pervasive
Ultimately, a company's investment in data warehouse and BI technology is only fully leveraged when BI becomes pervasive across the enterprise. This can only occur when BI-based access to the warehouse is as simple as it is powerful.

Dr. Joe Corn, a social and cultural historian at Stanford University, uses the automobile as a model for how technologies such as BI must evolve to ensure adoption. Driving the first cars required skill in lubricating various moving parts, adjusting spark plugs, setting the choke, opening the throttle, wielding the crank and knowing what to do when the car broke down. Drivers today simply turn the ignition key, put their foot on the accelerator, brake and steer. Drivers today are shielded from the complexity of modern cars—which are in fact sophisticated computers—by a greatly simplified and standardized interface.

By bringing this same combination of simplicity and sophistication to the analytical querying of data warehouses, next-generation BI is offering today's businesses the same benefits as modern drivers enjoy: the ability to go wherever they want to go. With the actionable information that BI provides, corporate decision-makers can drive financial and operational improvements, keep their organizations focused on clear goals, accurately model "what if" scenarios and cultivate a culture of accountability. The proven results of such transformation are optimized overall business performance and substantial competitive advantage. T

Continental scores with Teradata and Hyperion

Continental Airlines, the world's sixth largest airline, was recently named the "most admired global airline" by Fortune magazine. But Continental wasn't always so highly acclaimed. In 1994, it ranked 10th out of 10 airlines assessed by U.S. Department of Transportation metrics. The airline knew little about its important customers, set fares and schedules using conventional industry assumptions, conducted contract negotiations blindly and fought fraud only after the damage was done. Continental's turn around dramatically demonstrates how a data warehouse implementation and strategic use of BI—in this case, Teradata and Hyperion, respectively—can enable a company to attain competitive advantage. In fact, Continental's earned it Gartner's 2005 BI Excellence Award. Continental developed an enterprise data model that simplifies the joining of different subject areas to provide a single view of information for the BI environment. The model can support any query a user asks. Twenty-seven source systems feed the data warehouse, including schedules, inventory, reservations, OnePass (Continental's frequent flyer program) and employee/crew payroll. More than 1,100 people use Continental's Teradata/Hyperion-based system. The single, integrated, trusted view of the business has produced benefits ranging from better, faster decision-making to more than $500 million in cost savings, as well as incremental revenue from many initiatives that required BI information. The CRM and marketing team at Continental offers a conservative estimate of $150 million in revenue gains and $25 million in cost savings and fraud prevention.

(Click to enlarge) Effective visual presentation can greatly increase the benefits of BI. In the graphic above, the number of complaints generated by each age group is indicated by the length of the colored bar—while the profitability of each group is represented by its thickness. This analysis reveals that, of the customers likely to churn, those in the weekend caller segment are the most profitable.