Radiating IntelligenceBy David Baum
Oracle business intelligence turns information into intelligence and delivers it to the enterprise.
In a new vision for business intelligence (BI), all employees in an organization become potential decision-makers, whether they work in customer service, shipping, manufacturing, finance, human resources, or just about any other department. Companies with this vision are delivering analytical capabilities that are structured by use of a set of uniform definitions that can be recognized and applied by IT systems across the entire organization. This "pervasive BI" delivers information to business users in familiar ways and applications, and it's not just those familiar monthly and quarterly reports.
For example, financial services organization TIAA-CREF is rolling out a pervasive BI environment based on Oracle BI applications and the Oracle Business Intelligence Suite Enterprise Edition platform. "BI is not just about reporting," says Kurt Zimmer, vice president of CRM Delivery at TIAA-CREF. "It's the front door to how our business operates."
As one of the world's largest retirement systems, with more than US$380 billion in combined assets under management, TIAA-CREF employs 1,500 call center agents to serve 3.2 million individual investors and 600 field agents to serve its institutional investors. Oracle BI applications make the TIAA-CREF workforce more effective by giving agents greater insight into investor needs.
"The better we understand our investors, the better we can supply them with valuable services and help them meet their retirement goals," says Zimmer. "We're now looking at how we can use BI metrics to drive specific actions, such as alerting a representative when a customer request has not been dealt with in a set length of time."
TIAA-CREF began by using Siebel Business Analytics, which is now an integral part of the Oracle BI applications and Oracle Business Intelligence Suite Enterprise Edition, to track and analyze key contact center metrics such as first-call resolution rates, average call handle times, transfer rates, revenue per agent, and cost per contact. Now it also plans to use Oracle BI software to determine why people are calling and which issues merit immediate attention. The company's BI solution will enable employees to accumulate, analyze, and use customer data at the point of contact. A combination of workflow and analytic technology will help agents track open issues—such as when an investor requests a 401(k) loan—to make sure they are handled in a timely fashion.
"In many cases, these are time-critical requests, so we always want to be able to report on their status," Zimmer says. "As our analytical database grows, our service organization is becoming more proactive, by contacting customers to address suspected issues before they become problems, for instance."
Spreading the Analytical Wealth
TIAA-CREF's experience marks a departure from how BI has been deployed over the last two decades. In addition to giving managers front-end tools for query, reporting, and analysis, the company is embedding BI capabilities within its customer relationship management (CRM) applications to improve front-line sales, support, and case management activities. It is also creating a BI environment that helps call center agents automatically detect specific situations, notify key individuals to take action, or trigger a workflow that influences a related business process.
"Most people don't have time to search for problems and opportunities by looking through daily reports," explains Zimmer. "That's why we're embedding BI technology in the processes that our employees use all the time. Our strategy involves delivering BI information at the point of contact—ideally at the exact moment that it's needed."
Aydin Gencler, senior manager of product management and strategy at Oracle, uses the term insight-driven action to describe this type of business intelligence. The objective is to provide a unified view of the enterprise and enable business insight in an integrated environment—where BI and business applications are tightly integrated—to guide users down a path of discovery and action, he explains.
"Rather than forcing users to review long reports or query millions of records, a next- generation BI system should bring information into each user's daily business context to highlight opportunities and help that person effectively troubleshoot problems," he says. "That way, they will receive timely and relevant information based on triggers they set in advance and be guided to take action in response to business events."
Sometimes simple embedded BI functions whet users' appetites for more-advanced analytic capabilities. That's why Danske Spil A/S, the Danish National Lottery, has devised multiple methods for analyzing business data. "We're broadening access to our data warehouse and making it easier for users to generate reports on their own," says Margrethe Løjmar, a data architect and BI project manager at the firm.
Danske Spil has exclusive rights to sell games in Denmark to raise money for "good causes" designated by the Danish government. Its game portfolio includes many types of interactive, betting, and slot machine games distributed through the internet, retail chains, and cell phones. To help the company boost revenue and get to know its customers better, Løjmar and 15 other employees in the IT and business departments worked throughout 2006 on a new BI portal, which they launched in January 2007. As part of this process, they implemented Oracle Database 10g, Oracle Warehouse Builder 10g, and Oracle Business Intelligence Suite Standard Edition (including Oracle BI Discoverer) as the foundation for the new BI environment. Over time, Danske Spil plans to create customized data marts tied to departmental scorecards and dashboards and to offer data mining tools for advanced users.
"We're tailoring our BI platform to business users, not just professional analysts," Løjmar says. "We've modeled the warehouse so changes can be implemented quickly. Our new data model will decrease information overload and make it possible to create just-in-time information."
In a previous Danske Spil data warehouse, the level of aggregation was too detailed and the accompanying decision support tools were not suitable for most users. Løjmar says the adoption rate was low because of poor response times and a lack of user-friendly reporting tools. Both of these issues have been resolved in its new BI environment, which was completely revamped on the basis of user suggestions. "Letting business users drive your BI projects is the best approach," Løjmar says. "IT people often lose sight of what end users really need for making decisions."
Danske Spil learned another important lesson along the way: BI initiatives won't gain traction if users can't depend on the data. Thus, the Danish company is focused on improving data reliability and removing ambiguity from business terms and definitions. According to Løjmar, these efforts make users more productive and spare the IT department from having to constantly analyze and retrieve information, because end users have a good idea of what's in the database.
Banking on Quality
It was these same data management goals that motivated Absa Group Limited to create a consolidated BI environment that includes 1,000 reports and 31 BI projects. Absa is one of South Africa's largest financial services organizations, with US$66 billion in assets. According to Arthur Britz, general manager, Information Management Development, the company wanted a more cohesive BI strategy based on a standardized data model, methodology, architecture, tools, and measurement practices. "Our objective is not only to facilitate better decision-making, but also to drive actions that improve business performance," he says.
To get there, Absa decided to create an enterprise data warehouse by using Oracle Database 10g, Oracle Real Application Clusters (Oracle RAC), and Oracle Application Server 10g. The banking giant processes between 100 and 120 million records per day to create a repository of about 6 terabytes of analytical data for 4,000 users. Absa uses Oracle Warehouse Builder to gather source data from 43 core banking systems, as well as from third-party datasources, on a daily, weekly, and monthly basis. A unified data model ensures that managers throughout the organization obtain consistent answers from the data warehouse. "Oracle Warehouse Builder allows us to consolidate a multitude of sources into a single repository," Britz says. "In addition to loading billions of records each month, it provides us with auditability and traceability, which is essential for regulatory and compliance projects such as anti-money-laundering efforts, BASEL, and Sarbanes-Oxley."
Absa is implementing a BI methodology to enforce consistent strategic planning across the enterprise. End users rely on Oracle Business Intelligence Suite Standard Edition and Oracle Balanced Scorecard, in conjunction with third-party OLAP tools, to analyze BI data. Absa anticipates cost savings of more than 300 percent over five years as a result of its consolidation efforts.
Making It Work
A pervasive BI environment requires a pervasive software infrastructure. That's why Absa depends on Oracle RAC, Oracle Data Guard, and Oracle Automatic Storage Management (Oracle ASM) to ensure that its BI environment is always available to its large and active user base, which usually includes about 300 concurrent users.
"Instead of making them focus on operational management issues, Oracle Data Guard and Oracle ASM let our DBAs focus on strategic development and administration functions," says Britz. "Along with Oracle RAC, these software products help us meet the service-level expectations of our users and respond to rapid growth in our business intelligence and data warehousing solutions."
Having a highly available BI solution driven by comprehensive data management practices becomes progressively more important as BI capabilities are extended to users throughout the organization. TIAA-CREF is taking a tiered approach with its data model based on each user's role within the organization. "We reorganize information at every layer to make sure it is relevant for the individuals using it," says TIAA-CREF's Zimmer. For example, agents see information about their own day-to-day activities and operations managers can view summaries of an entire division of the call center. At the highest level are dashboards for senior managers that summarize key performance metrics and alert them when certain thresholds are reached or critical service levels are exceeded.
"We are continually collecting data about our customers: their products, their holdings, their transactions, their demographic information, and so forth," says Zimmer. "The beauty of Oracle BI is that we can map this very complex structure to logically represent the way our business looks. It allows us to take a very complex data model and describe our business in simple terms. That knowledge is embedded in our systems, and the analytic technology reveals it to people in an actionable way."
Assembling the Pieces
Oracle's Gencler describes Oracle's BI platform as "hot-pluggable," which means that it is designed to work with heterogeneous environments integrating data from multiple relational, unstructured, OLAP, and prepackaged application sources. "Customers can protect their existing investments by easily integrating their existing applications, tools, and data warehouse environments with Oracle's BI platform," he says. As these BI platforms evolve, the technology is used to analyze data generated by operational processes, rather than simply to report on historical data.
"Many people are finding that they can benefit greatly from a real-time view of business events, such as identifying high-value sales orders or flagging discrepancies in a bill of materials," says Henry Morris, group vice president and general manager for the Integration, Development, and Application Strategies (IDeAS) solutions research group at market intelligence firm IDC. In conjunction with business activity monitoring software, BI systems can analyze performance data, prompt managers to take corrective action, and provide alerts when certain conditions are met, such as when a threshold is exceeded in a supply chain planning system.
This is where the Oracle BI platform benefits from the other components in the Oracle Fusion Middleware family, such as business process orchestration from Oracle BPEL Process Manager. Customers can use Oracle BPEL Process Manager to embed BI functions as part of a workflow: to run reports, execute database queries, send e-mail, update records, and so forth.
Morris says a complete BI platform typically includes pervasive end-user capabilities along with integrated tools for security, collaboration, content management, event monitoring, data quality, and related middleware functions. "This makes it easier to integrate new BI capabilities with existing enterprise applications and to create business processes that mimic the way people work," he says. In this context, BI capabilities commonly take the form of services that are invoked though a service-oriented architecture, enabling them to be easily displayed through portals, dashboards, and business processes.
Danske Spil is achieving this vision with an Oracle-based infrastructure for storing, managing, and analyzing data, accessed through a BI portal that provides a workspace for sharing information and amassing business knowledge. This integrated online environment addresses a broad spectrum of analytical requirements to help people make decisions based on current facts and conditions.
"By closely monitoring product performance, our managers can determine which gaming services are the most profitable and make instant adjustments to their promotion strategies," Danske Spil's Løjmar says. They can also more easily monitor the effect of advertising campaigns on revenue, as they get closer to a just-in-time BI environment.
Predicting the Future
"Helping people make decisions based on real-time conditions is the starting point for shifting from reactive to proactive modes of thinking," says IDC's Morris, offering such examples as monitoring crime patterns and analyzing product defects on an assembly line. By identifying problems and sending alerts accompanied by relevant information to appropriate systems or individuals, companies can quickly spot opportunities and resolve anomalies before they become problems.
TIAA-CREF's Zimmer believes that this type of real-time foresight can improve activities in a call center as well. "Historically, we've had groups of individuals that capture information at month-end so we can analyze the results," he says.
"Looking in the rearview mirror gives us a good view of what's happened, but increasingly we're also interested in predicting what's ahead, so we can anticipate events and take action," he says. It's not just about sales metrics or service metrics. It's also about cause and effect. When you can connect marketing, service, and sales, you start to see relationships that weren't obvious before."
That's exactly how BI technology is influencing new types of users, activities, and business processes. "BI is no longer a standalone activity," Zimmer says. "It's integrated into how we do business."
David Baum (email@example.com) is a freelance business writer based in Santa Barbara, California.