Companies faced with ever-expanding information sources both in and outside the firewall are scrambling to find easier ways to collect, aggregate, and report on data and make it available throughout the enterprise.
At first blush, that would seem to be the territory of database reporting tools, and to some extent it is, except that most corporate information does not reside in structured databases. In fact, a great deal of it is unstructured information such as documents or email or, even harder to pin down, it's somewhere out on the World Wide Web waiting to be found. Perhaps it's in a newsfeed or it can be found in a business database such as Dun & Bradstreet, but wherever it is, it is important for companies to be able to make use of and share with others in the enterprise.
This type of information-gathering falls within the broad category of software known as "business intelligence" tools, which allow employees to learn about a particular subject such as how the sales department is performing, what customers complain about most or what your closest competitor is doing. This article looks at the range of business intelligence software and provides some examples of how companies are using this information to run their businesses more efficiently.
The BI Landscape
Since Business Intelligence (BI) spans a wide range of information types, it's not an easy area to nail down. For some, it means looking at pure business information such as sales data broken down by territory and pulled directly from a database. For others, it's less structured information such as internal PowerPoint presentations or other business documents or information found on the Web. Databases and data warehouses provide a large source of business information, but they are far from the only sources available to the enterprise. Actually, unstructured data found in places other than databases comprise the vast majority of data in the enterprise. "Today about 15 percent of the data are in a structured form and 85 percent of the data is in unstructured form. There is a big industry around that 15 percent," says Anant Jhingran, director of business intelligence at IBM.
According to Dan Vesset, research manager, analytics and data warehousing at IDC in Framingham, Massachusetts, that 15 percent is part of a huge industry IDC defines as business analytics (BA), which includes query and reporting tools, multi-dimensional analysis tools, datamining, and packaged data marts. Vesset says that BA accounted for 12 billion dollars worth of business in 2002. IDC sees BI as a piece of the broader business analytics market, which Vesset says accounted for 3.7 billion dollars of the total business analytics pie.
Beyond the data found in conventional databases, there is a whole area of business information sometimes referred to as marketing intelligence (from CRM sources, for example), business intelligence, or competitive intelligence. This type of information might be found inside the firewall or out on the open Web. Employees need to not only find this information, but make it available to their fellow employees. John Blossom, president of Shore Communications, Inc., a content industry research firm, sees portals and knowledge management tools playing an important role in gathering and distributing this information. Blossom says, "Portal software and knowledge management is fairly key to success in business intelligence. Being able to transmit business intelligence into the organization is a key factor, not only to collect it, but to get it actionable. So you see major corporations, not only gathering information to be distributed in reports, but making it available online."
Whichever information type or source companies use, a tool to help extract the data exists. In the structured data market, that might be Cognos or Business Objects. In the unstructured market, you might use a visual taxonomy tool such as Inxight SmartDiscovery to get a grip on the data in your enterprise (or from the Web), or an unstructured data search tool such as Insightful's InFact. You might look outside the firewall with tools such as IBM's nascent WebFountain product or Anacubis Desktop, which gives you a front end to help you make sense of data you find on Web-based subscription services such as D&B. Let's look at these options more closely now.
Making Sense of Structured Information
One thing is certain: There is no shortage of information sitting in enterprise databases today. The trouble comes when companies try to actually make sense of that data. Throughout the 90s, many companies spent loads of money and time building huge data warehouses. Today, they want to use that information they have gathered to build a competitive advantage. "Now the business units, the business people in the organization are saying, ‘OK we have those data warehouses and we need to get some information out of there, make better decision, drive our business better, help us run our business.' BI feeds off databases and presents information to business users in an organization in a way that's meaningful." says Anil Dilwari, product marketing manager at Cognos, an enterprise BI software vendor.
Dilwari says his company helps organize information by displaying data in a visual window. Cognos presents this information in a single interface, but breaks down the gathering process into a process that includes what they call a scorecards, dashboards, OLAP (online vertical processing technology) analysis, reporting, and event detection. By providing a smooth path through each of these different functions, Cognos software gives users a way to get the big picture and drill down to find the answers they need.
Business Objects is another company trying to help companies make sense of structured data. Their strength lies in giving the end-user control of the query process. Early on, that meant giving end-users the power to write their own queries without IT assistance and more recently providing a set of tools in an integrated end-user interface. Darren Cunningham, group manager for data integration at Business Objects says, "Our early value proposition was for a patented technology that we call a semantic layer, which essentially shields end-users from having to work with any underlying programming language, from having to write their own queries, really giving end-users the ability to analyze the data that lives in these disparate systems." Today, they offer a full set of data analysis tools.