Business Intelligence By Clay Stephens, Brian Peterson, & Kosta Golemis BA 370 Dr. René F. Reitsma 11/15/2005 Abstract: This paper discusses the topic of business intelligence, and touches on multiple areas within this subject. Areas discussed include a history and the definition of business intelligence, how we gather it, a discussion of data warehouses and data marts, the business intelligence data warehouse (BIDW) process, business and government uses of business intelligence, and the technology used to gather it. 1 Introduction Business intelligence (BI) is a relatively new thing to hit the business world. Only recently did it start to garner attention, being introduced in the early 90’s with lots of promise. Since then, BI has had its ups and downs, but is currently a point of focus for business managers and IT professionals. Gathering information has always been a high priority for businesses (no matter what field they may be in), and business intelligence has greatly affected the way businesses gather information and what information they can obtain. It is no wonder that business intelligence has attracted so much interest in the business world, as all companies can benefit from it, and is now almost a necessity if for no other reason than to keep up with competitors. What is Business Intelligence? The term business intelligence refers to the process of taking raw data, turning those data into information, and then making business decisions on the basis of that information. An example would be a grocery store recording every sale it makes in a day, being able to print out a report that states how many of a particular product has been sold in a given time frame, and from that make a decision on when to put this product on sale. This has potential to make and save companies lots of money, as it makes decision making easier and can provide valuable information that had otherwise been ignored. History Business intelligence first started to make an impact in the business environment in the early 90’s. Computer technology started to really come about around then and people started talking about how business intelligence was going to save companies millions (or perhaps even billions) of dollars. BI had a big surge in popularity, and everyone started recording market data into databases. Towards the late 90’s, however, the surge of BI slowed to a standstill. Companies had started the process, but didn’t know how to finish it because nobody really knew how to get from gathering information to saving money. Also, the vision of BI wasn’t supported by current technology. Processor speed was a lot slower back then, and it took a lot longer to record these data into databases and extract them, thus hampering the process. Tools that help this process were not widely available back then as either. Recently, however, business intelligence has made a big comeback. Technology is now up to par with the requirements of efficiently extracting data, and there is a wide variety of tools to help companies with the data extraction process. People have had a few years to discuss how this can help save money, and BI is now being used with those goals in mind and companies have more realistic expectations out of it. 2 Gathering Business Intelligence The process of gathering business intelligence starts with data integration. This is the process of combining databases into one consolidated database for sharing and analysis. After this is complete, a process called data mining can start. Data mining is the process of extracting previously unknown and useful information from large data sets or databases. The mined data are then stored in a data warehouse or a data mart. A data warehouse is a special database where all the useful information is stored for easy access. A data mart is essentially the same thing, but with more specific information of predetermined selected data. This process of extracting the data into a smaller database makes for more efficient data analysis and reporting. After this is complete, a company must use these data to help it make business decisions such as planning and forecasting. This is the most important step, as not taking any action would render all the previous steps useless. The BIDW Process The business intelligence data warehousing (BIDW) process (see figure 1) begins with data integration. Part of data integration includes extraction, transformation, and loading (E.T.L.). Extraction is the phase where data are extracted from outside sources, converting the data into fields. In the transformation phase a series of rules or functions are applied to the extracted data, making the data fit to what the business needs. Loading the data into the data warehouses is the last phase of the E.T.L. process. Figure 1: BIDW Process Source: Business Intelligence and Data Warehousing The raw data are stored, retrieved, and updated by an organization’s on-line transaction processing (OLTP) system. The OLTP is a system for collecting and managing the base data in an organization. An ATM machine or an electronic employee time clock is an example of an OLTP system. None of the data collected have analytical value. The information is then cleansed (duplicate information is removed, for example) and optimized, and data are structured for queries and analysis. In the next two phases of the BIDW process, data mining, query, and analytical tools generate intelligence which enables companies to spot trends, enhance business relationships, and create new opportunities. This intelligence is used to make decisions, create strategies, and implement programs for competitive advantage. 3 Meanwhile, throughout the whole process the system is regulated by an overall corporate security policy. A company wants a strong security policy because the data collected and analyzed are typically confidential and critical to a business’ operations, therefore all the data must be secure from external and internal threats. In the end, if a company wants to have a well-run BIDW operation it needs to have business performance management (BPM) applications, which help track the results of the decisions made and the performance of the programs created. Uses of Business Intelligence Business Intelligence uses include government, retail, financial services, and Telco (see figure 2). Business Intelligence can be modified to work for a variety of companies and agencies, but in the end it can be separated into two main fields: government and business. Figure 2 (brief breakdown of service sector for BI, Source: www.computerworld.com) According to ComputerWorld there are three factors that shape BI today. 1) Today's BI software is mostly reactive; focused on transactions and static data. 2) In five years, BI will move beyond the boundaries of an organization and expand into the supply chain. This means that BI will become better at recognizing and understanding customer’s behavior. And lastly, with security becoming a major priority for enterprises, 3) Companies will adopt an information-aware infrastructure model in the next five years that will support global, distributed BI while automatically protecting the security and privacy of information (ComputerWorld, The Future of BI, 19 Sept 2005). “Over the next three years, the pervasiveness of business intelligence technology will explode, by as much as fivefold, as three technology trends converge: the availability of 64-bit in-memory processing technology improving BI performance by factors of 10 to 100; the service-enabling of business applications; and the growth of rich Internet applications, or RIA. Together, they enable organizations to leverage BI in new and innovative ways. Imagine you've been driving with the parking brake on, and suddenly you release it. You not only go cheaper and faster to the same places as before, but you 4 also explore new places and opportunities.” -- Lothar Schubert, director of SAP NetWeaver product marketing, Walldorf, Germany (ComputerWorld, The Future of BI, 19 Sept 2005) BI applications will become much more commonplace using applications such as spreadsheets. BI applications themselves are also becoming much easier to use so that non statisticians and non IT individuals will be able to use them. A dashboard is a layer within an operating system that hosts mini applications. Figure 3 is an example of what a dashboard for the common end user would be using for BI. Figure 3 (BI dashboards built from Oracle’s Database, Source: http://www.oracle.com/technology/pub/images/rittman_dash_f1.gif) Business Uses Businesses use business intelligence to analyze competitors, conduct market and industry research, and try to gain the competitive advantage. Take one such example: You own a retail store and customer Joe buys a silk shirt. Now when the customer is rung up, his information is stored into the computer. It turns out that Joe buys four silk shirts every six months- with BI a report can be sent to the manager stating that Joe has a pattern of buying items that are alike. The manager now has the knowledge about these patterns and he could send out a silk shirt flier in the future to Joe. This could possibly 5 get Joe to buy more shirts than he normally does with knowledge of sales/events from these fliers. Government Government use of BI includes sharing data between local and federal intelligence agencies to reduce crime levels and lower the risk of terrorist attacks. It also helps further develop a uniform system of government BI resources, in which local, state, and federal services are available online (such as the DMV, social security, etc.). One hypothetical example is as follows: A rural neighborhood has had six unsolved burglaries in the last five months. BI will look for patterns that people may not think to look for. For instance it may turn out that seven months ago an ex-con moved into an apartment building less than half a mile away, and that he was originally in jail for burglary. Take this real government agency example: North Carolina Department of Health and Human Services is responsible for ensuring the health, safety, and well being of 7 million North Carolinians. This department has nearly 20 divisions with over 19,000 employees. They try and share information across departments and keep track of records on millions of individuals. If the agency did not complete required reports on time they risked loosing federal funding and/or paying large fines. They also needed a tool that could help case workers view important information about clients more easily. They decided to go with a data warehouse system and chose BI tools from Business Objects to deliver critical information to its end users in easy-to-understand reports. Employees in the divisions are now able to correlate data from other case outcomes. This allows staff to identify higher level trends and improve service to their customers. They also use the new reporting system to run clean-up queries that identify cases coded incorrectly in the data warehouse. By identifying these coding mistakes faster than before, they were able to save $1 million per year on administrative costs that would have otherwise been spent on fixing these errors. With less time and money spent on paperwork they are able to focus on serving the needs of their clients. Software in Business Intelligence There are many software firms that produce BI tools and software, ranging from Oracle & Microsoft to even Sun Microsystems (see figure 4) - although the company that seems to be leading them all is Cognos. Cognos is an Ottawa, Ontario based company which makes BI and performance planning software. They were founded in 1969 and employ 3,300 people and serve more than 23,000 customers in over 135 countries. Cognos has continually demonstrated return on investments to its customers and partners. Cognos has been recognized by Network Computing Magazine as Editor’s Choice for leading BI intelligence solutions. Cognos Metrics Manager was awarded Intelligent Enterprise Magazine’s Readers’ Choice award for best business monitoring solution, & Cognos was also named one of the Top 12 Most Influential Enterprise Software Companies by Intelligent Enterprise Magazine. (Cognos) 6 Figure 4 (Direction of BI software companies. Source: unknown) Lee Martin believes that in the next 6-7 years that of the very large number of BI products that exist today there will only be about five left. This will be due to mergers, buyouts, bankruptcies and other companies that just plain outperform others. Conclusion Business Intelligence has a lot of practical use, and can save companies a lot of money. However, it will not solve all the problems in a company, and one must have the correct tools in order to properly extract business intelligence. The process can be time consuming, but the outcome can be very useful and beneficial in the long run. Several software companies are currently developing BI software, but in the future we may see some change in that BI may become even more popular, and that a few companies may overtake the smaller companies to dominate the market. We will have to wait and see what business intelligence brings us in the future. 7 References .ComputerWorld (10 Nov. 2005) The Future of BI; Available: http://www.computerworld.com/databasetopics/businessintelligence/story/0,10801,10458 9,00.html Cognos (10 Nov. 2005) Awards; Available: http://www.cognos.com/news/awards/index.html Wikipedia Encyclopedia (8 November 2005), Extract, transform, load;. Available: http://en.wikipedia.org/wiki/Extract%2C_transform%2C_load Sun Microsystems (2005), Business Intelligence and Data Warehousing; Available: http://www.sun.com/storage/white-papers/bidw.pdf 8