Operational BI: Getting “Real-Time” about Performance
A new survey sheds light on how enterprises are using and benefiting from operational BI
by David Hatch
Between October 15, 2007 and December 15, 2007, Aberdeen Group examined the use, the experiences, and the intentions of more than 250 enterprises involved in “operational BI” initiatives and projects.
Aberdeen supplemented this online survey effort with 30-40 telephone interviews with select respondents, gathering additional information on operational BI strategies, experiences, and results. Responding executives completed an online survey that included questions designed to determine:
- The degree to which operational BI is deployed within current operations and the financial implications of the technology
- The structure and effectiveness of existing implementations
- Current and planned use of operational BI to aid activities across a broad range of business activity and processes
- The benefits, if any, that have been derived from operational BI initiatives
The study sought to identify emerging best practices for operational BI usage, and to provide a framework by which readers could assess their own management capabilities. Aberdeen used five key performance criteria to distinguish what we classify as Best-in-Class (BIC) from Industry Average and Laggard organizations:
- Decrease in time-to-information
- Decrease in time-to-decision or time-to-action
- Customer satisfaction Composite Index Result (CIR), measured as the percentage improvement in customer satisfaction, issue resolution speed, and accuracy
- Percentage improvement in customer retention
- Percentage improvement in system data access and availability for end-users
Companies are trying to improve efficiencies and performance of many day-to-day and real-time activities, such as customer interactions, finance and accounting processes, transportation/shipping, sales activity, manufacturing, inventory management, business activity monitoring (BAM), and marketing. Aberdeen research conducted for its report, “Smart Decisions: The Role of Key Performance Indicators” (September, 2007), found that Best-in-Class companies are improving their time-to-decision by implementing capabilities, technologies, and services that enable faster delivery of mission-critical information to the people who need it, when they need it, and how they need it. This report reveals the current and planned initiatives that companies favor to improve operational efficiencies and timeliness of actions.
The "Flavors" of Operational BI
The term "operational BI" has been identified with several terms within various industry and market vernaculars. Some of the terms have been coined to differentiate between the timeframes within which data collection, reporting, and analysis occur. Other terms describe differences in the methods and calculations that take place as data is captured, manipulated, and delivered.
The "flavors" of operational BI include:
- Transactional BI: Analysis and reporting capabilities embedded within transactional systems, run as needed or daily
- Real-time analytics: Automated analysis generated from business rules and algorithms applied to data as it is captured, with second or sub-second response
- Near-real-time analytics: Automated analysis generated from business rules and algorithms applied to data as it is captured, run every second to hourly
- Operational reporting: Reports automatically generated (and often distributed) based on business rules, algorithms, or raw data as it is captured from transactional or integrated data sources, run at any user-specified interval
- BAM or business process monitoring: Typically refers to the monitoring of systems or process performance and activity, including reporting, analytics, and automated actions based on predetermined targets; run every second, minute, or in real time
- Decision management: Rules-based engines integrated with reporting and analytic applications that are designed to automate actions based on exceptions, thresholds, and other rules, run at any user-specified interval
No matter what flavor or definition is used, at the heart of the matter, organizations are beginning to focus their attention on leveraging existing data to enable and optimize daily, hourly, minute-to-minute, or even up-to-the-second actions.
Operational BI is not necessarily a single reporting or analytics application applied to a specific set of data. It involves the ability to access information affecting the business as the data is created. This can include one or a multiple set of data sources, and can affect one or many sets of decisions, actions, and people. Organizations that take a strategic approach to operational BI, and the access to relevant data when, how, and where people need it, will be better positioned to achieve Best-in-Class success. This requires the ability to dynamically collect and integrate data and make it available to decision makers for improvement of time-to-decision, or to business rules engines and management systems for automation of actions. There is an added degree of difficulty as data volumes and complexity continues to grow.
While traditional business intelligence solutions continue to address the strategic information needs of decision makers with analysis of historical data, organizations are starting to realize the potential of applying BI technology and approaches to “dynamic information” – data that changes multiple times during the business day, hour, or minute that requires knowledge workers to make immediate decisions.
Key benefits to end users include opportunities to:
- Manage business activities as they occur, as opposed to waiting for the end of a day, week, month, or project timeline before gaining access to analytical data and information
- Improve customer relations and reduce risk by responding to events more rapidly, and heading off harmful events before they occur or the customer becomes aware
- Increase business efficiency by providing actionable information to line-level knowledge workers in real time, and automating manual processes to reduce costly, repetitive report-creation tasks
Best-in-Class organizations -- the companies achieving the highest performance based on survey metrics -- have shown that "time-to-decision" (the interval between actual business activity and action or decision) is a key performance metric driving success with operational performance improvement initiatives. 63 percent of Best-in-Class companies have improved their time-to-decision within the past 12 months, compared to just 23 percent of Industry Average companies and only 6 percent of Laggards.
Figure 1: Best-in-Class KPI - Percentage of Companies Reporting
Decrease in Time-to-Decision/Action within the Past 12 months
Source: Aberdeen Group, December 2007
Aberdeen's research survey included additional granularity about the degree to which organizations have achieved improvement in time-to-decision. Figure 1 (above) illustrates the gap between Best-in-Class companies and Industry Average and Laggard organizations. While most companies have not achieved second-to-second timeframes, Best-in-Class companies have improved their time-to-decision from days-to-hours and from hours-to-minutes.
Traditional BI initiatives are driven at a department or project level, no matter what level of maturity a company has achieved, as cited in Aberdeen’s September 2007 research brief, Enterprise BI - Comparing the BI Giants. Operational BI, however, is more apt to start as a strategic enterprise initiative among Best-in-Class companies (51 percent) compared to Industry Average companies (33 percent). A strategic enterprise-level view of operational information alleviates some of the barriers organizations face as they attempt to create logical views of information that are integrated from multiple data sources across a diverse set of departmental systems.
Process: Let the Data Do the Work
Best-in-Class companies are more likely to automate their operational data processing and analysis activities. By letting the data do the work, Best-in-Class organizations are able to arrive at key insights faster, deliver critical information to the right people at the right time, and (in some cases) produce analysis required for rapid decisions and actions. Many of the companies interviewed are starting to automate actions based on changing operational data. This capability is becoming prevalent within financial services organizations where fluctuations in currency, stock valuations, and key index metrics happen at second and sub-second intervals. Manufacturers are also beginning to deploy technologies that automatically analyze production, yield, waste, and quality control metrics.
Performance and Risk Management
Operational BI initiatives offer the ability to detect performance changes automatically, and therefore take corrective action before real harm is done. This is particularly important in customer-facing activity. Companies that can detect and act upon performance changes (such as customer satisfaction and response time) are therefore able to identify and solve problems, often before the customer is aware of them.
How an organization measures its operational performance has a lot to do with whether or not it achieves Best-in-Class status. Sixty-five percent of Best-in-Class respondents measure and track operational performance against departmental or enterprise goals. Only half of Industry Average companies, and just over a third of Laggards do the same.
When asked about the nature of the operational BI technology and service enablers currently in use, Best-in-Class companies reported a significant advantage within several categories. Of the top five enabling technologies and services, operational forecasting was rated highest. This is important in understanding how Best-in-Class companies are approaching operational BI projects. It is not just the access, calculation, and delivery of data that is important. The ability to forecast operational performance pertains to collecting operational data and using it to predict short-term results. For example, marketers are learning to forecast operational results of online campaigns by utilizing real-time and near-real-time data as it becomes available from current campaign activity.
Best-in-Class organizations are planning to drop their reliance on static reporting and MS Excel in favor of a broad range of technology enablers. The planned technology investments, such as automated alert reporting and notification, align with planned capability investments. As companies strive for a faster time-to-information and time-to-decision, they are therefore also heavily investing in technologies and services that deliver enhanced automation capability. Forty-three percent of Best-in-Class companies plan to access operational BI capabilities that are embedded within enterprise applications. This indicates a continuing trend towards reliance on one technology provider for reporting and analytical tools.
Editor's note: The full report is available until 2/29/08 athttp://www.aberdeen.com/summary/report/benchmark/4522-RA-operation-bi-perf.asp. (Short registration required.)
David Hatch is research director of Aberdeen Group’s Business Intelligence Practice, where he benchmarks user organizations’ BI strategies, actions, and planned technology investments. His research focuses on the collection (DQ, EDI), assembly (data modeling, data access, business rules, and KPIs), and delivery (reporting, dashboards, scorecards, analytic tools, portals, On-Demad/SaaS, and software licensing) of information in the enterprise. David holds a BA, Communications degree from the University of Massachusetts. You can reach the author at firstname.lastname@example.org.