The problem: Too much data and not enough tools to analyze the data to make decisions. Solution one: Insert analysis software everywhere. Solution two: Present analysis results as scorecards, speedometers, traffic lights, performance maps, and other user-friendly, simple-to-read graphics. Therein is found the benefit of business intelligence (BI). Done right, BI can provide people with the right information at the right time. New BI products have a wealth of query and reporting tools and advanced multidimensional data visualizations. They are easy to use and customize. Knowledge about arcane BI mechanics, data warehousing, and structured query language are no longer required. And, "pervasive BI" is more a reality because the technologies for software integration are so much better.
The BI market dramatically changes
The BI market got a major recharge through recent market consolidations. IBM bought Cognos (Burlington, MA; www.cognos.com
; January 2008). Oracle Corporation bought Hyperion (Santa Clara, CA; www.hyperion.com
; April 2007). SAP (Newtown Square, PA; www.sap.com/sapbusinessobjects
) bought Business Objects (San Jose, CA; January 2008). Microsoft Corp. (Redmond, WA;) bought ProClarity (Boise, ID; April 2006). By the end of 2008, these four companies, the world's largest software vendors, controlled about half of the BI software market. Last, analysis software vendor SAS Institute Inc. bought Teragram (Cambridge, MA; www.teragram.com
; March 2008).
Each vendor's BI strategy is different. Cognos is crucial as the informational/analytical foundation for IBM's "Information On Demand" strategy. Plus, tightening the integration between Cognos and Lotus Notes (an earlier IBM acquisition) could produce more collaborative BI based on searching and analyzing unstructured enterprise data. For Microsoft, embedding BI tools inside all Microsoft desktop products, especially Office, should make data analysis and display easier, more sophisticated, and more accurate. For Oracle, Hyperion BI infuses new software capabilities to the company's vast product line of enterprise software. SAP claims that Business Objects will be "the new face of BI for SAP." A BI front-end can help SAP customers gain more insight into huge amounts of transactional enterprise data. (Also, because a large percentage of Business Objects's customers use Oracle applications and databases, SAP customers get the added advantage of complementing their SAP software with Oracle products, which themselves are now loaded with Hyperion BI capabilities.) Over at SAS, Teragram's natural language processing and advanced linguistic technology (basically, recognizing the meaning of text) helps turn "monstrous amounts" of data into usable information. This would be a natural fit for SAS analytics.
Pretty, and pretty informative
Pervasive BI requires integration techno-logies and strategies such as service-oriented architecture, simple object access protocol, Sun Java, Microsoft .NET, and eXtensible Markup Language. Much of BI is becoming web-based because the web itself is so pervasive. To support such BI activity, vendors have developed single, integrated BI platforms. For instance, IBM has its IBM Cognos 8 v4, Oracle its Oracle Business Intelligence Standard Edition One and Suite Enterprise Edition Plus, SAP its BusinessObjects Enterprise and BusinessObjects Edge (for midsize companies), SAS its Enterprise BI Server. These systems combine enterprise planning, analytics, and reporting into a single, centralized system. They address a variety of BI requirements, including ad hoc reporting and analysis, dashboards and visualization, data integration and live data access (even to competitive software applications, databases, data warehouses, and analyst tools), data cleansing, prepackaged data mart systems, web-based reporting and data querying. They provide "data lineage": letting users view the data in context within a BI report. These platforms can connect to multiple multi-vendor data sources in multiple remote locations. Last, these platforms are easy to deploy (read, faster and overall less expensive).
But that's all in the background. Much of the value of BI applications is in presenting information. For example, look at the IBM Cognos 8 Go! family of applications. The Go! Office module lets people view, interact with, and refresh data within Microsoft Office. The Go! Mobile module sends real-time reports, analysis, and alerts to wireless devices. These outputs automatically adjust to the location of the device. The Go! Dashboard adds pizzazz to presentations-operational, tactical, and strategic-while actually communicating relevant, accurate information. These presentations can include Adobe Flash gauges, maps, charts, and external widgets, such as RSS feeds or prebuilt reports, scorecards, and metrics. Users can drag and drop, re-sort, and rearrange dynamic graphical content to display multiple data, drill down into a report, create dashboards, and create web-based content with composite views-without needing IT to build individual dashboards for every business scenario.
The goal of pervasive BI is to embed BI tools everywhere, including portals and especially specific applications. All of these are matched to the user ("contextual BI"). They're also easy to use so that people don't realize they're using some very sophisticated software. A little BI can go a long way. For example, one of Oracle's "prebuilt" BI packages is its Agile PLM Business Intelligence product, which helps decision-making during new product introductions, specifically the design-to-release process. This product helps companies improve overall design for manufacturability, product quality, and cost while minimizing supply chain disruptions. Through role-based and functional dashboards, and reports, Agile PLM BI analyzes and displays key metrics and trends across the product quality management processes, focuses on resource utilization, and provides drill-downs and analysis to investigate potential product quality risks. The BI in Oracle Supply Chain and Order Management Analytics analyzes order and inventory data so that companies can assess inventory levels to determine product fulfillment needs before an order is booked, identify potential order backlog problems, manage accounts receivable and daily sales, and measure their effectiveness in managing raw materials and finished goods inventories across multiple locations. Event-based and scheduled alerts automatically highlight problems for users through email and handheld devices, and to their personalized dashboards.
SAS Service Parts Optimization forecasts parts demand, helping service organizations maintain adequate stock levels and optimize response times. This BI tool automatically generates millions of statistically based, multilevel forecasts at frequent intervals by gathering and consolidating large volumes of data, regardless of format and location. It will calculate optimal inventory levels and policies throughout the service chain based on user-specified constraints (such as service levels, lead times, and costs). It also incorporates all variables on both the supply side (such as inventory on hand, on order, committed, and in transit) and the demand side (such as forecasts and sales orders) for the entire network, as well as every single location.