It could be a scene from Martin Scorsese's film, Casino. The camera pans a room of baccarat tables. And there, we see the boss ... Robert De Niro?

Well, no. It's former Harvard Business School professor and MIT Ph.D., Gary Loveman, Harrah's COO. Instead of ruthless charm, Dr. Loveman relies on sophisticated analytical tools to retain customers and encourage further spending. Harrah's estimates that its customer loyalty program-which categorizes the millions of visitors to its 25 domestic properties into 64 different customer segments-is responsible for more than $100 million in incremental revenues. Harrah's also drills down into its databases to price hotel rooms on a real-time basis, factoring in anticipated profitability and room availability. This intensive data crunching is paying off. Revenues for the first nine months of 2002 were $3.1 billion, up from $2.7 billion in the same period a year before.

The foundation of Harrah's strategy is analytics: accessing, aggregating, and analyzing large amounts of data from diverse sources to understand historical performance or behavior, or to predict- or manage-outcomes. In short, Harrah's is converting data into knowledge that enhances its ability to make effective business decisions.

Harrah's adoption of analytics signals a recognition that a successful business will be one that understands its customers' behavior at a most granular level.

Incremental, but Real Progress

No one "great leap forward" has driven this new generation of analytics; instead, incremental developments are making a visible impact. Dramatic decreases in computing and data storage costs means most large firms can afford the sophisticated analyses. And today's data analysis tools are more advanced than those available even five years ago and run on pervasive operating systems such as UNIX and NT.

Moreover, companies are now capturing more information about their customers. Today, marketers that have electronic gateways-such as Internet portals, banking machines, telephones, or slot machines-can retrieve detailed demographic and behavioral information about individuals from third-party data providers in less than a second. And in the same period of time, they can harness this information for programmed decision making.

In fact, users of today's advanced analytic tools can pinpoint the most significant variables-among countless possibilities-and clearly identify interrelationships. Consider how the pharmaceuticals industry uses analytics. Pharmaceutical researchers often need to determine whether one molecular structure in huge databases of perhaps a million molecular structures is biologically active against a specific disease. The researchers can tailor software to accomplish this seemingly impossible analysis. The ability to analyze molecular structures and predict drug effectiveness has helped researchers discover new applications for existing drugs as well as develop new medications.

Five Rules of the Road

The foundation of data analytics has always been a company's existing data resources, which typically reside in a number of discrete legacy databases or data warehouses. Getting those discrete databases to "talk" to one another was nearly impossible with early data mining products, and a great deal of data reentry was needed to assemble all theinformation in one place. Today, powerful computers and sophisticated toolkits make it significantly easier to tap into all the databases at once.

But while tapping into various databases is now easier, analytics has become more complex. Today's systems store progressively larger amounts of data, and they store more types of data:

More data doesn't have to mean greater confusion. The systems and data are accessible-it's how you leverage the data and transform it into valuable information that makes all the difference. Here are five rules of the road to ensure that an organization's analytics implementation leads to the ability to make better business decisions.