In a nutshell, a business intelligence platform extracts valuable data from diverse business systems, stores it in a consistent format, provides the tools to analyze it for hidden and not-so-hidden trends and relationships, and helps you report your findings in ways that non-technical users can understand. The principal components of a business intelligence platform are the data warehouse, data mining and analysis tools, and reporting tools-each examined below.

The data warehouse is a kind of staging area that stores and primes data for analysis. Corporate data is generally spread across numerous relational database, e-commerce, Web, ERP, CRM, and other systems, each of which has its own database architectures, data models, errors, inconsistencies, and missing information. It's almost impossible to analyze data that's so diverse and inconsistent, and manipulating original data can be a dangerous proposition. So a data warehouse prepares data for analysis by performing four essential functions:

Data warehousing functionality can be found in many different types of software. It may be built into or available as an add-on to a relational database management system (RDBMS) such as Microsoft SQL Server, IBM DB2, or Oracle 9i. You can also find products that specialize in BI and data warehousing such as SAS, SPSS Base, or Hyperion Essbase.

In many cases, data warehousing may be just the first step in preparing data for analysis. Large amounts of heterogeneous data can be made more manageable and responsive by distilling data warehouses even further into a number of individual "data marts," or filtered data stores that are subject and application specific. For example, you may pull data out of a data warehouse to build separate sales and financial data marts.

Once data has been extracted, cleansed, distilled, and stored appropriately, it's ready for analysis. Business intelligence platforms provide a virtual cornucopia of analysis tools that can expose relationships, patterns, and trends across departments, business functions, and markets. You can use these tools to focus on specific information and relationships to support highly targeted decision making, or you can drill deep down to reveal hidden relationships you may never have considered before. For example, retail operations can perform complex analyses to determine which products sell best at what price, in what combinations, at what locations, with how much shelf space, and with what types of promotions. Hospitals can drill into mountains of historical data to determine exactly how different complex combinations of treatments affect outcomes in different types of patients.

One of the most valuable tools here is an Online Analytical Processing (OLAP) engine, which can examine multidimensional data from many different angles to find numerous hidden relationships. OLAP engines are built for high scalability and performance, so they can service thousands of users with multiple queries and complex calculations in seconds. OLAP enables users to move beyond historical information to predict future trends and build complicated "what if?" scenarios that can be approached from many different perspectives and business assumptions. For example, a user can determine in seconds what the affect of a five percent increase in advertising outlays may be on the sales of a particular product line, or what affect certain arrangements of products in a warehouse may have on worker accidents.

There are also tools that allow you to work backwards, segmenting and grouping customers in multiple different ways in order to produce a desired outcome. OLAP can also be used to provide your e-commerce sites with valuable personalization features that analyze purchase patterns in order to recommend appropriate product purchases to individual customers. You also get scads of statistical tools and functions such as such as linear regression, factor analysis, hierarchical cluster analysis, and many more. These are the tools you use to find new ways to improve your company's revenue and profitability.

Once you have finished the complex data analysis, you'll want to produce useful reports, often for a non-technical audience. Most BI platforms make it relatively easy for you to produce concise, attractive, readable reports and present them in the form of Web pages, Word documents, PowerPoint presentations, XML data, or any number of different presentation formats. Most offer different chart, table, and graph types, and other useful graphical tools. Many can even produce interactive graphics for users who want to look at data from multiple dimensions, rotate charts, and perform their own data manipulation to see how results are affected.

While there are comprehensive business intelligence platforms that incorporate most or all of the functions outlined above, there are other products that tackle only one or two functions, so you can build your own platform using components you may already have, and then fill in the missing pieces with new solutions. Tools that tackle just a single BI niche, such as Web site log-file analysis, are also available; as well as a range of tools that targets everything from the entry-level small business user to the large enterprise.

SPSS is an example of a comprehensive, integrated, modular business intelligence platform that you can pick and choose from according to your needs. You'd usually start with SPSS Base for its data warehousing functionality, mainstream analysis tools, and reporting capabilities, and add more specialized analysis modules from there if you need them. The nice thing about such a solution is that you don't have to worry about how easily the numerous pieces will work together, or duplicate work you've already done to fit the requirements of another product by another vendor.

On the other hand, you may already be familiar with and prefer a best-of-breed product such as Crystal Decisions Crystal Reports. This solution excels in rich reporting capabilities using content from numerous data sources, and comes with many powerful and easy-to-use report wizards as well as some OLAP functionality. You can use Crystal Reports or Crystal Enterprise to produce many different types of Web-based reports, and even export them to XML and PDF formats if you choose. Crystal Reports also has excellent security features that allow you to restrict views of certain data, or allow some users to manipulate data and others only to view it. If Web site analysis is your main concern, NetIQ, WebTrends Analysis Suite, which comes in Standard and Advanced editions, provides numerous tools that small and medium-sized businesses can use to analyze Web site activity, functionality, and effectiveness. Among other things, it can analyze your log files to deliver valuable information on the activity of visitors to your site, including how they got there, the pages where they spent the most time, and where they exited. For more general-purpose, entry-level users there's Microsoft Data Analyzer, which adds business intelligence to Microsoft Office XP, including analysis and visualization capabilities that complement those of Excel and other Office components. Data Analyzer is also designed to optimize some of the advanced features of Microsoft SQL Server 2000 Analysis Services.

Knowledge is one of your most valuable strategic assets. By taking advantage of business intelligence tools, you can gain the insight that you need quickly, in order to thrive in an increasingly competitive e-business environment.