|
Operational business intelligence investments are only as good as the data they deliver, and unfortunately, many enterprises believe that business intelligence (BI) begins with data in a database or warehouse and ends with a management report. However, scalable, effective BI systems optimize data access and processes at the source ? the vital information systems that run the business. When properly constructed, operational business intelligence systems provide critical input and feedback to everyone in the enterprise and beyond ? not just to managers and business analysts.
A successful operational system is built on a business intelligence data integration framework that enables an organization to create a repeatable, consistent approach for acquiring, transforming, integrating, and delivering data to applications and end users who need it. Information Builders'WebFOCUS provides a powerful integration framework combined with advanced features to place highly effective business intelligence within reach of most organizations. By maximizing data at all of its sources, WebFOCUS helps organizations improve the effectiveness of their existing BI tools as well as enhance the quality of the information they use for driving business operations.
In most people's minds, business intelligence begins with a data warehouse or datamart and ends with an online analytical processing (OLAP) tool or report. However, the data itself doesn't originate in a repository and most of an organization's data never ends up in one. Data actually originates in operational transaction systems, packaged and legacy applications, individual's spreadsheets, and from sources external to the organization. These origins are invisible to the traditional BI user or implementer, who have no sense of the time, cost, and complexity involved in gathering and preparing disparate data to make it available and useful.
There is a better way to understand business intelligence. When examined as an end-to-end process, traditionally BI has relied upon an elaborate infrastructure comprised of these transaction systems, applications, multiple sources, electronic data exchanges, and many disparate, often labor-intensive processes. These processes must access, gather, reconcile, and prepare data to deliver an accurate view of the business to BI users.
For example, data for the sales analysis application of a large automobile manufacturer originates in Excel spreadsheets, forwarded from all over the globe. Each spreadsheet's contents and structure differs depending on the country of origin. All of these diverse spreadsheets must be reformatted, reconciled, and loaded, by hand, into a data store so that the data is available for monthly sales analysis processes.
Recognizing Access and Integration Issues
Today's business intelligence needs often demand that source data be available in real time, as well as for historical analysis. They also demand that the resulting information be consistent and accurate. Therefore, accessing data for use in BI actually begins where source data is created or captured ? in operational transaction systems, packaged applications, and legacy applications, as well as electronic business processes through which businesses exchange data. Data access is critically important because the more robust the data-access capabilities of a BI system, the more easily, cost-effectively, and consistently information may be produced and delivered to all of the information consumers who need it.
Yet as Gartner points out, data integration issues consume a significant majority of the effort expended in a BI project. "Designing a repeatable process by which data is acquired from operational systems, transformed, integrated, and delivered to the data warehouse is technically challenging." All of these data origins store, represent, and exchange data differently, typically in mediums and formats that are incompatible with easy access, representation, or manipulation by business intelligence consumers.
Establishing a BI Integration Framework
The ability of BI tools to exploit individual source interfaces is critical to accessing quality data from its original sources. Sometimes applications need variable-length virtual storage access method (VSAM) records or hierarchical data in IMS databases or extensible markup language (XML) files. Sometimes integration requires access to systems through messaging interfaces like WebSphere MQ. Sometimes data isn't directly accessible at all, and an application must access stored procedures, API calls, or Web services instead. Rather than seeking a single silver bullet to resolve a wide range of integration challenges, organizations are better served by establishing a unified BI data integration framework.
|
